As filed with the U.S. Securities and Exchange Commission on May 20, 2026

Registration No. 333-    

UNITED STATES

SECURITIES AND EXCHANGE COMMISSION

WASHINGTON, DC 20549

FORM S-1

REGISTRATION STATEMENT

UNDER THE SECURITIES ACT OF 1933

Space Exploration Technologies Corp.

(Exact name of registrant as specified in its charter)

Texas	7370	01-0627671
(State or other jurisdiction of incorporation or organization)	(Primary Standard Industrial Classification Code Number)	(I.R.S. Employer Identification Number)

1 Rocket Road

Starbase, Texas 78521

(Address, including zip code, and telephone number, including area code, of registrant’s principal executive offices)

Elon Musk

Chief Executive Officer

1 Rocket Road

Starbase, Texas 78521

Tel: (310) 363-6000

(Name, address, including zip code, and telephone number, including area code, of agent for service)

With copies to:

George J. Sampas Hillary H. Holmes Harrison Tucker Atma J. Kabad Gibson, Dunn & Crutcher LLP 811 Main Street, Suite 3000 Houston, Texas 77002 Tel: (346) 718-6600 Bret Johnsen Michael Smith Space Exploration Technologies Corp. 1 Rocket Road Hawthorne, California 90250 Tel: (310) 363-6000 Byron B. Rooney Alan F. Denenberg Stephen A. Byeff Joze Vranicar Davis Polk & Wardwell LLP 450 Lexington Avenue New York, New York 10017 Tel: (212) 450-4000

Approximate date of commencement of proposed sale to the public:

As soon as practicable after this Registration Statement becomes effective.

If any of the securities being registered on this Form are to be offered on a delayed or continuous basis pursuant to Rule 415 under the Securities Act of 1933 check the following

box.  ☐

If this Form is filed to register additional securities for an offering pursuant to Rule 462(b) under the Securities Act, check the following box and list the Securities Act registration

statement number of the earlier effective registration statement for the same offering.  ☐

If this Form is a post-effective amendment filed pursuant to Rule 462(c) under the Securities Act, check the following box and list the Securities Act registration statement number

of the earlier effective registration statement for the same offering.  ☐

If this Form is a post-effective amendment filed pursuant to Rule 462(d) under the Securities Act, check the following box and list the Securities Act registration statement number

of the earlier effective registration statement for the same offering.  ☐

Indicate by check mark whether the registrant is a large accelerated filer, an accelerated filer, a non-accelerated filer, a smaller reporting company or an emerging growth company.

See the definitions of “large accelerated filer,” “accelerated filer,” “smaller reporting company” and “emerging growth company” in Rule 12b-2 of the Exchange Act:

Large accelerated filer	☐	Accelerated filer	☐
Non-accelerated filer	☒	Smaller reporting company	☐
		Emerging growth company	☐

If an emerging growth company, indicate by check mark if the registrant has elected not to use the extended transition period for complying with any new or revised financial

accounting standards provided pursuant to Section 7(a)(2)(B) of the Securities Act.  ☐

The Registrant hereby amends this Registration Statement on such date or dates as may be necessary to delay its effective date until the registrant shall file a further

amendment which specifically states that this Registration Statement shall thereafter become effective in accordance with Section 8(a) of the Securities Act of 1933, as

amended, or until the Registration Statement shall become effective on such date as the Securities and Exchange Commission, acting pursuant to such Section 8(a), may

determine.

SUBJECT TO COMPLETION, DATED      , 2026

PRELIMINARY PROSPECTUS

Shares

Space Exploration Technologies Corp.

Class A Common Stock

This is the initial public offering of shares of Class A common stock, par value $0.001 per share, of Space Exploration Technologies

Corp., a Texas corporation. We are offering                shares of our Class A common stock.

Currently, no public market exists for our Class A common stock. We expect the initial public offering price to be between $    and

$    per share. We have applied to list our Class A common stock on The Nasdaq Stock Market LLC (“Nasdaq”) and Nasdaq Texas,

Inc. (“Nasdaq Texas”) under the symbol “SPCX.”

Following the completion of this offering, we will have two classes of common stock issued and outstanding: Class A common stock

and Class B common stock. Each share of Class A common stock will entitle its holder to one vote per share. Each share of Class B

common stock will entitle its holder to 10 votes per share. Class A shareholders and Class B shareholders will vote together as a

single class on all matters to be voted on by shareholders, except Class B shareholders will be entitled to elect a majority of our board

of directors in addition to having certain other class votes as described under “Description of Capital Stock.”

Assuming an offering size as set forth above and an initial public offering price of $                per share (the midpoint of the estimated

price range set forth above), Elon Musk, our founder, Chief Executive Officer, Chief Technical Officer and Chairman of our board,

will hold approximately           % of the voting power of our common stock (or approximately        % if the underwriters exercise their

option to purchase additional shares of Class A common stock in full) immediately after the completion of this offering through his

ownership of shares of our Class A and Class B common stock of which approximately           % he controls through his ownership of

our Class B common stock. As a result, Mr. Musk will be able to control the outcome of matters requiring shareholder approval. This

includes the election of (i) a majority of our board, through his ownership of Class B shares (as Class B Directors), for so long as he

holds a majority of the voting power of the Class B common stock, and (ii) the remainder of our board, for so long as he holds a

majority of the combined voting power of the Class A and Class B common stock. As a result, we will be a “controlled company”

under the corporate governance rules of Nasdaq following the completion of this offering and, as a result, we intend to rely on

exemptions from certain corporate governance requirements. Please refer to “Management—Controlled Company Exemption.”

Investing in our Class A common stock involves risks. Please refer to “Risk Factors” beginning on page 26 of this

prospectus.

The information in this preliminary prospectus is not complete and may be changed. The securities described herein may not be sold until the registration statement filed with the Securities and Exchange

Commission is effective. This prospectus is not an offer to sell such securities, and it is not soliciting an offer to buy these securities, in any jurisdiction where the offer or sale is not permitted.

	Per Share	Total
Initial public offering price......................................................................................................	$	$
Underwriting discounts and commissions(1)............................................................................	$	$
Proceeds, before expenses, to Space Exploration Technologies Corp....................................	$	$

________________

(1)Please refer to “Underwriting” for a description of all underwriting compensation payable in connection with this offering.

The underwriters may also exercise an option to purchase up to an additional       shares of our Class A common stock from us, at the

initial public offering price, less the underwriting discounts and commissions, for 30 days after the date of this prospectus.

At our request, the underwriters have reserved up to             percent of the shares of Class A common stock to be issued by the

Company and offered by this prospectus for sale, at the initial public offering price, to              . Please refer to “Underwriting—

Directed Share Program.” Neither the Securities and Exchange Commission (the “SEC”) nor any state securities commission has

approved or disapproved of these securities or passed on the adequacy or accuracy of this prospectus. Any representation to the

contrary is a criminal offense.

The shares of Class A common stock will be ready for delivery on or about             , 2026.

Joint Book-Running Managers

Goldman Sachs & Co. LLC Morgan Stanley BofA Securities Citigroup J.P. Morgan

Barclays Deutsche Bank Securities RBC Capital Markets UBS Investment Bank Wells Fargo Securities

Allen & Company LLC Cantor Needham & Company Raymond James Societe Generale Stifel William Blair

BTG Pactual ING Macquarie Capital Mirae Asset Securities Mizuho Santander

Prospectus Dated              , 2026.

 TABLE OF CONTENTS

	Page
GLOSSARY OF TERMS.................................................................................................................................	iv
PROSPECTUS SUMMARY............................................................................................................................	1
RISK FACTORS..............................................................................................................................................	26
CAUTIONARY STATEMENT REGARDING FORWARD-LOOKING STATEMENTS...........................	64
USE OF PROCEEDS.......................................................................................................................................	66
DIVIDEND POLICY........................................................................................................................................	67
CAPITALIZATION.........................................................................................................................................	68
DILUTION.......................................................................................................................................................	70
MANAGEMENT’S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND RESULTS OF OPERATIONS........................................................................................................................................	74
BUSINESS........................................................................................................................................................	130
MANAGEMENT..............................................................................................................................................	226
EXECUTIVE COMPENSATION....................................................................................................................	233
CERTAIN RELATIONSHIPS AND RELATED PERSON TRANSACTIONS.............................................	243
SECURITY OWNERSHIP OF CERTAIN BENEFICIAL OWNERS AND MANAGEMENT....................	247
DESCRIPTION OF CAPITAL STOCK..........................................................................................................	250
SHARES ELIGIBLE FOR FUTURE SALE....................................................................................................	258
MATERIAL U.S. FEDERAL INCOME TAX CONSIDERATIONS FOR NON-U.S. HOLDERS OF CLASS A COMMON STOCK.....................................................................................................................	260
UNDERWRITING...........................................................................................................................................	264
LEGAL MATTERS..........................................................................................................................................	277
EXPERTS.........................................................................................................................................................	277
WHERE YOU CAN FIND ADDITIONAL INFORMATION........................................................................	277
INDEX TO FINANCIAL STATEMENTS......................................................................................................	F-1

Neither we nor the underwriters have authorized anyone to provide you with information other than that contained in

this prospectus or in any free writing prospectus authorized by us. We and the underwriters take no responsibility

for, and can provide no assurance as to the reliability of, any other information that others may give you. We and the

underwriters are not making an offer to sell, or seeking offers to buy, our Class A common stock in any jurisdiction

where an offer or sale is not permitted. The information contained in this prospectus or any free writing prospectus is

accurate only as of its date, regardless of its time of delivery or of any sale of shares of our Class A common stock.

Our business, financial condition, results of operations and future prospects may have changed since that date.

For investors outside of the United States: Neither we nor the underwriters have done anything that would permit

this offering, or possession or distribution of this prospectus, in any jurisdiction where action for that purpose is

required, other than the United States. Persons outside of the United States who come into possession of this

prospectus must inform themselves about, and observe any restrictions relating to, the offering of the shares of our

Class A common stock and the distribution of this prospectus outside of the United States.

This prospectus contains forward-looking statements that are subject to a number of risks and uncertainties, many of

which are beyond our control. Please refer to “Risk Factors” and “Cautionary Statement Regarding Forward-

Looking Statements.”

ii

General Information

Except as otherwise indicated or required by the context, all references to “SpaceX,” the “Company,” “we,” “our”

and “us” or similar terms refer to Space Exploration Technologies Corp. and its consolidated subsidiaries. For the

definitions of certain terms and abbreviations used in this prospectus, please refer to “Glossary of Terms” beginning

on page iv of this prospectus.

References to (i) our “bylaws” are to the form of amended and restated bylaws of the Company (as amended and

restated from time to time) to be effective upon the completion of this offering, (ii) our “charter” are to the form of

restated certificate of formation of the Company to be effective upon the completion of this offering and (iii) “our

board” or “the board” are to the board of directors of the Company.

Basis of Presentation

The consolidated financial statements of SpaceX have been retrospectively recast for all periods presented to include

(i) the historical results of X.AI Holdings Corp., which was acquired by SpaceX, effective February 2, 2026 (the

“xAI Merger”), and X Holdings Corp. (“X Holdings”), which was acquired by xAI, effective March 28, 2025 (the

“X Merger”), because these transactions were between entities under common control, and (ii) a five-for-one stock

split of the Company’s Class A, Class B, and Class C Common Stock, effective May 4, 2026 (the “2026 Stock

Split”). Unless otherwise stated or the context otherwise requires, all share and per share information included in this

prospectus have been retroactively adjusted to reflect the 2026 Stock Split. Refer to Note 1, Nature of Business, to

the audited consolidated financial statements included elsewhere in this prospectus.

Industry and Market Data

Certain market and industry data and forecasts used in this prospectus have been obtained from, are based on, or use

data from, the following reports and sources, among others: (i) Breaking Barriers to Data Center Growth, dated

January 20, 2025, by Boston Consulting Group; (ii) Looming Spectrum Shortfall Could Cost America’s GDP $1.4T,

Jeopardize Continued Function of U.S. Networks, New Report Finds, dated March 27, 2025, by the Cellular

Telecommunications and Internet Association; (iii) Top 50 Countries by Number of Business Aircraft Registered,

dated January 27, 2026, by Corporate Jet Investor; (iv) Digital Economy Trends 2026, dated December 2025, by the

Digital Cooperation Organization; (v) Global Fixed Broadband Market Outlook, Ericsson Mobility Report, dated

November 1, 2025, by Ericsson; (vi) Households by Number of Households and by Country, Euromonitor

International Passport 2026 Edition, dated November 5, 2025, by Euromonitor International; (vii) Satellite Solutions

for Universal Service, dated March 2025, by the Global Satellite Operators Association; (viii) Broadband Services

Market Analysis Segment Forecast to 2027, dated April 2025, by Grand View Research; (ix) Consumer Market

Model H2 2025 – Worldwide Household Internet Penetration, dated March 2026, by International Data Corporation;

(x) World Energy Outlook Special Report: Energy and AI, dated April 2025, by the International Energy Agency;

(xi) The 175 GW Crisis: America’s Power Grid Cannot Keep Up with AI Data Centers, dated January 21, 2026, by

Introl; (xii) As Wireless Network Quality Competition Increases, Customers Benefit, dated July 17, 2025, by J.D.

Power; (xiii) Satellite Statistics: Satellite and Debris Population, dated April 2026, by Jonathan McDowell; (xiv)

2026 Global Data Center Outlook: Navigating AI Demand, Power Constraints and Global Opportunities, dated

January 5, 2026, by JLL; (xv) Global Ship Tracking Intelligence, at marinetraffic.com, as updated from time to time

and last accessed April 13, 2026, by Marine Traffic Dashboard; (xvi) The Cost of Compute: A $7 Trillion Race to

Scale Data Centers, dated April 28, 2025, by McKinsey & Company; (xvii) What is Multimodal AI?, dated June 10,

2025, by McKinsey & Company; (xviii) NASA: Enabling America on the Space Frontier, dated December 2024, by

the National Aeronautics and Space Administration (“NASA”); (xix) Space Act Agreement, dated April 2015, by

NASA; (xx) The Recent Large Reduction in Space Launch Cost, dated July 8, 2018, by NASA; (xxi) 12th Edition

Space Economy Report, dated January 29, 2026, by Novaspace; (xxii) Global Fleet and MRO Market Forecast

2025–2035, dated February 2025, by Oliver Wyman; (xxiii) Broadband Op Subs by Technology – Forecasts

Summary, dated March 31, 2026, by Omdia; (xxiv) Mobile Forecasts Summary – February 2026, dated February

18, 2026, by Omdia; (xxv) Data Center Rules and Regulations, dated September 8, 2025, by QTS; (xxvi) AI’s

Power Requirements Under Exponential Growth, dated January 28, 2025, by RAND Corporation; (xxvii) Data

Center Grid-Power Demand to Rise 22% in 2025, Nearly Triple by 2030, dated October 14, 2025, by S&P Global

Market Intelligence; (xxviii) NVIDIA GTC 2025 – Built for Reasoning, Vera Rubin, Kyber, CPO, Dynamo

iii

Inference, Jensen Math, Feynman, dated March 18, 2025, by SemiAnalysis; (xxix) NVIDIA Blackwell Ultra

Datasheet, dated February 16, 2026, by SemiAnalysis; (xxx) H100 Rental Price Over Time (2023–2025): A

Complete Market Analysis, dated December 21, 2025, by Silicon Data; (xxxi) Data Centers – Understanding the

Power Consumption of Data Centers, at socomec.us, as updated from time to time and last accessed April 13, 2026,

by Socomec; (xxxii) The Space Report 2025 Q2 Highlights Record $613 Billion Global Space Economy for 2024,

dated July 22, 2025, by the Space Foundation; (xxxiii) Median Country Speeds Updated February 2026, dated

February 2026, by the Speedtest Global Index; (xxxiv) Data Center (Russian Market) Commercial Data Centers,

dated January 28, 2026, by TAdviser; (xxxv) Merchant Fleet by Flag of Registration and by Type of Ship, dated

June 10, 2025, by the United Nations Conference on Trade and Development; (xxxvi) U.S. Electricity Generation in

2025 Hit a Record, Again, dated March 5, 2026, by the U.S. Energy Information Administration; (xxxvii)

GAO-25-107555, In-Space Servicing, Assembly, and Manufacturing: Benefits, Challenges, and Policy Options,

dated July 2025, by the U.S. Government Accountability Office; (xxxviii) GDP (current US$), at

data.worldbank.data.org, as updated from time to time and last accessed April 13, 2026, by the World Bank; (xxxix)

Rural population (% of total population), at data.worldbank.org, as updated from time to time and last accessed May

2, 2026, by the World Bank; (xl) How Data Centres in Space Sustainably Enable the AI Revolution, dated January

16, 2026, by Philip Johnston Co-Founder and Chief Executive Officer, Starcloud, published by the World Economic

Forum; and (xli) Most Americans Use AI but Still Don’t Trust It, dated December 9, 2025, by YouGov. We did not

commission the preparation of any of these reports or sources.

Some market data and statistical information contained in this prospectus are also based on management’s estimates

and calculations, which are derived from our review and interpretation of publicly available industry publications,

our internal research and our knowledge of the markets in which we currently, and will in the future, operate, as well

as the sources referred to above. This information involves a number of assumptions and limitations, and you are

cautioned not to give undue weight to such information. The estimates and assumptions used in determining our

total addressable markets are further detailed in the section titled “Business—Our Market Opportunity,” and you are

urged to read the risk factor titled “The estimates of future market opportunity and forecasts of market growth, and

our ability to capture such markets, included in this prospectus may prove to be inaccurate.” Forecasts and other

forward-looking information obtained from the sources named above are subject to the same qualifications and

uncertainties as the other forward-looking statements in this prospectus.

Statements as to market position, market opportunity and market size are based on data currently available to us, as

well as management’s estimates, judgments, assessments, and assumptions. While we are not aware of any

misstatements regarding market position, market opportunity, and market size information included in this

prospectus, such information, which is derived in part from management’s estimates and beliefs, is inherently

uncertain and imprecise. Projections, assumptions and estimates of estimated market position and market

opportunity and the future performance of the industries in which we operate are necessarily subject to a high degree

of uncertainty and risk due to a variety of factors, including those described in “Risk Factors,” “Cautionary

Statement Regarding Forward-Looking Statements” and elsewhere in this prospectus. These and other factors could

cause results to differ materially from those expressed in the estimates made by third parties and by us. Investors are

cautioned not to place undue reliance on statements of expected future market size or opportunity.

Trademarks and Trade Names

We own or have rights to various trademarks, service marks and trade names that we use in connection with the

operation of our business. This prospectus may also contain trademarks, service marks and trade names of third

parties, which are the property of their respective owners. Our use or display of third parties’ trademarks, service

marks, trade names or products in this prospectus is not intended to, and does not imply, a relationship with us or an

endorsement or sponsorship by or of us. Solely for convenience, the trademarks, service marks and trade names

referred to in this prospectus may appear without the ®, ™ or SM symbols, but such references are not intended to

indicate, in any way, that we will not assert, to the fullest extent under applicable law, our rights or the right of the

applicable licensor to these trademarks, service marks and trade names.

iv

GLOSSARY OF TERMS

The terms and abbreviations defined in this section are used throughout this prospectus:

•“AI” or “artificial intelligence” refers to advanced computational technologies and systems enabling machines

to learn, comprehend reality, solve complex problems, exhibit creativity, make critical decisions, and function

with growing autonomy.

•“AI compute” or “compute” refers to the computing infrastructure required to train and operate artificial

intelligence models, including, without limitation, specialized processors, networking, storage, and power

systems deployed in data centers or other computing environments.

•“AI compute satellite” refers to a satellite equipped with onboard artificial intelligence processing capabilities

designed to perform data analysis, inference, or other machine learning, automated decision-making and

artificial intelligence algorithms, models and technologies workloads in orbit.

•“AI ecosystem” refers to a complex, multi-layered network of technologies, products, systems, and

infrastructure that develop, leverage, and deploy intelligent systems.

•“AI segment” refers to our AI business, which we acquired in connection with our acquisition of xAI in

February 2026, and includes our AI compute, Grok, and X.

•“AI training cluster” refers to an integrated system that provides computational power required for training and

running advanced AI models.

•“The Algorithm” refers to our five-step iterative process that we use to rapidly innovate and optimize,

emphasizing making the requirements less dumb, deleting unnecessary processes or parts, optimizing the

necessary processes or parts, accelerating cycle timesteps, and automating only proven processes after the first

four steps are completed.

•“Application Programming Interface” or “API” refers to a defined set of rules and protocols that allows

different software systems to communicate with and interact with each other programmatically.

•“ARPU” refers to service revenue generated from Starlink Subscribers during a period divided by (i) the

average number of Starlink Subscribers during the period and by (ii) the number of months in the period.

•“Artemis program” refers to a NASA program aimed at landing humans on the Moon by the late 2020s.

•“booster” refers to the first-stage rocket that provides the primary thrust during launch.

•“booster catch” refers to a recovery method in which a returning first-stage rocket booster is captured mid-air by

mechanical arms on the launch tower rather than on legs at a landing zone or at sea.

•“booster launch” refers to a rocket launch in which a booster stage provides the primary thrust during liftoff and

the initial phase of ascent before separating from the vehicle.

•“bps” refers to bits per second.

•“COLOSSUS” refers to our flagship data center, located on Paul R. Lowry Road in Memphis, Tennessee.

•“COLOSSUS II” refers to our data centers in Memphis, Tennessee and in Southaven, Mississippi. These data

centers are part of our coherent gigawatt-scale AI training cluster.

•“Connectivity segment” refers to our Connectivity segment, which includes Starlink and associated offerings.

•“Credit Agreements” refers to our SpaceX Credit Facility and SpaceX Bridge Loan.

•“crewmember” refers to a person who has traveled on our spacecraft, measuring by each mission.

v

•“daily posts” on X and Grok refers to the aggregate volume of original posts, replies, reposts, quotes and media

shared daily by users on the X platform, and the real-time interactions, analysis and generative capabilities

provided to a user by Grok. This may include posts generated by AI or accounts managed by AI.

•“downlink capacity” refers to the maximum rate at which data can be transmitted from a satellite to users over a

network or communication link in a given period of time.

•“Draco thrusters” refers to thrusters used in Dragon spacecraft for precise orbital maneuvering and adjustments.

•“Dragon” refers to our Dragon spacecraft.

•“Falcon 1” refers to our two-stage, liquid-fueled small-lift launch vehicle that operated from 2006 to 2009.

•“Falcon 9” refers to our orbital-class rocket with reusable boosters, first launched in 2010, which has a payload

capacity to LEO of approximately 23 metric tons.

•“Falcon Heavy” refers to our partially reusable super heavy-lift launch vehicle, first launched in 2018, which

has a payload capacity to LEO of approximately 64 metric tons.

•“flight-proven booster launches” refers to a mission utilizing a booster that has previously completed at least

one successful launch and recovery.

•“frontier model” refers to a leading-edge, sophisticated large language model, such as Grok, designed for

rigorous reasoning and real-time information synthesis.

•“Gbps” refers to gigabits per second.

•“geostationary orbit” refers to a high Earth orbit that allows satellites to match Earth’s rotation, appearing

stationary from the ground, often used for communication satellites.

•“geosynchronous transfer orbit” refers to an elliptical orbit used to transfer a spacecraft from a lower orbit to a

geostationary orbit.

•“gigawatt” refers to one billion watts.

•“gigawatt-scale” refers to infrastructure, systems, or facilities that are designed to generate, transmit, or

consume approximately one gigawatt or more of electrical power capacity.

•“GPU” refers to a graphics processing unit.

•“Grok” refers to our family of frontier models, which represents a core pillar of our mission to advance

humanity’s understanding of the universe through the development of truth-seeking artificial intelligence.

•“Grok API” refers to our application programming interface that enables developers to access and integrate

Grok models into external software applications and workflows.

•“Grok Business” refers to our subscription-based offering that provides organizations with access to Grok

models and related tools for use in internal business applications and workflows, designed for deployment by

small-to-medium teams.

•“Grok Enterprise” refers to our subscription-based offering that provides organizations with access to Grok

models and related tools for use in internal business applications and workflows, designed for deployment by

enterprise organizations.

•“Grok Voice” refers to the Grok real-time speech engine.

•“high-density compute” refers to compute infrastructure designed to deliver a large amount of processing power

within a limited physical footprint, typically characterized by high processor concentration and elevated power

usage per unit of space.

vi

•“Imagine” refers to our image and video generation system.

•“inference” refers to the process by which a trained artificial intelligence model generates outputs (such as text,

images, or predictions) from new input data.

•“International Docking System Standard” refers to a standard for autonomous docking capabilities used by

spacecraft like Dragon.

•“IoT” refers to the network of physical objects embedded with sensors, software, and other technologies for the

purpose of connecting and exchanging data with other devices and systems over the internet.

•“Kardashev Type II” refers to a civilization that harnesses the full energy output of its local star, like our Sun, to

power unprecedented growth and sustain the civilization’s existence.

•“large language model” or “LLM” refers to a sophisticated artificial intelligence model designed for advanced

reasoning and natural language processing.

•“large-scale LEO broadband satellite constellation” refers to a satellite constellation network of over 1,000

satellites.

•“latency” refers to the time delay between the transmission of data from a source and its receipt at a destination,

typically measured in milliseconds.

•“launch payload mass” refers to the theoretical payload mass that a particular spacecraft is capable of delivering

to a specified orbit under specific conditions, which is derived from advanced computer simulations and

performance modeling that apply to particular mission scenarios and trajectory assumptions. Actual payload

that can be delivered for a given mission may be different and will vary depending on numerous mission

parameters and operational factors, including mission-specific trajectory requirements, atmospheric conditions,

vehicle and payload configuration, risk profile, and applicable regulatory or range-safety limitations.

•“launch system” refers to a comprehensive system comprising rockets and associated ground infrastructure used

to launch spacecraft and payloads into space.

•“launch vehicle” refers to a rocket designed to transport payloads from terrestrial bodies (e.g., Earth, Moon, or

Mars) to space or to a designated orbital trajectory.

•“LEO satellite constellation” refers to a network of numerous satellites operating in Low-Earth Orbit, typically

deployed to provide services such as broadband connectivity, including Starlink.

•“Low-Earth Orbit” or “LEO” refers to an orbit relatively close to Earth’s surface, typically used by satellites for

applications like broadband internet due to its lower latency compared to higher orbits.

•“low-latency network” refers to a network with latency below 70 milliseconds.

•“lunar mass driver” refers to a launch system that we intend to build on the Moon’s surface that will be

designed to use electromagnetic acceleration to propel payloads into space without the use of rockets.

•“Macrohard” refers to a platform we are currently developing that is designed to emulate digital workflows,

augment human operation of computers, and create a fully AI-operated software company.

•“mass to orbit” refers to the total kilograms of payload deployed to orbit in a given period, and is a key indicator

of our capacity and scalability that supports Space revenue and drives expansion across our Connectivity and AI

segments.

•“MAU” (or monthly active users) refers to the total number of users who have interacted with Grok or X

through web browsers or mobile applications at least once during the 30-day period ending on the date of

measurement (“active users”). In presenting combined MAUs across the two platforms, we seek to identify and

account for users who access both Grok and X based on sign-in traffic so that such users are not double-counted

vii

when measuring MAU. Furthermore, only users who have registered for an X or Grok account are included.

While we believe our methodologies provide a reasonable approximation of MAU based on the number of

unique users, they may not fully capture all instances of duplication, and our reported MAU should be viewed

as an estimate of unique users across our Grok and X platforms for the applicable period. We track the subset of

users who used Grok’s AI features and those who have not based on the source of their server requests.

•“Mbps” refers to megabits per second.

•“Megapack” refers to a containerized, utility-scale lithium-ion battery energy storage system produced by Tesla

and designed to stabilize power grids, store renewable energy, and replace fossil fuel peaker plants.

•“megawatt” refers to one million watts.

•“Merlin” refers to the Merlin family of engines, which include vacuum and sea level variants and are fully

developed and produced by the Company.

•“microgravity” refers to very weak gravity, such as that experienced in orbiting spacecraft, which allows for

unique manufacturing processes like creating ultra-pure materials.

•“Mid-Earth Orbit” or “MEO” refers to an orbital region between approximately 2,000 km and 35,786 km above

Earth’s surface.

•“mission success rate” refers to the proportion of Falcon 9 and Falcon Heavy missions that achieve their

primary objectives. This term does not include Starship flight tests.

•“mobile network operators” or “MNOs” refers to the local entities of the companies that provide mobile phone

services to customers, with whom SpaceX partners to offer satellite-to-mobile connectivity. The term may also

include mobile virtual network operators, where applicable.

•“Mobile Satellite Service” refers to providing wireless voice, messaging, and data connectivity to, from, or

between mobile devices by using orbiting satellites rather than terrestrial cell towers.

•“Moore’s Law” refers to an observation, not a physical law, that the number of transistors on a microchip

doubles roughly every two years, leading to exponentially faster, smaller, and cheaper electronics.

•“orbital AI compute” refers to artificial intelligence computing infrastructure contemplated to be deployed in

space, consisting of satellite constellations that act as orbital data centers, harnessing solar energy for power and

leveraging the space environment for cooling. We expect to begin deploying our orbital AI compute satellites as

early as 2028.

•“payload” refers to the portion of a vehicle’s total mass that consists of the cargo, passengers, satellites, or other

mission-specific items being transported and that reaches the target orbit or destination. Payload is distinct from

total mass (also referred to as gross mass or initial mass) which is the entire weight of the vehicle, including the

payload, fuel / propellant, structure, engines, and any other items, at the start of a journey.

•“payload capacity to orbit” refers to a theoretical payload capacity that a particular launch vehicle is capable of

delivering to a specified orbit (e.g., LEO or GEO) or celestial body (e.g., Mars) under specific conditions, which

orbit is derived from advanced computer simulations and performance modelling that apply to particular

mission scenarios and trajectory assumptions. Actual payload capacity for a given mission may be different and

will vary depending on numerous mission parameters and operational factors, including mission-specific

trajectory requirements, atmospheric conditions, vehicle and payload configuration, risk profile, and applicable

regulatory or range-safety limitations.

•“Power Usage Effectiveness” refers to the global standard metric for data center efficiency, calculated as the

ratio of total facility power to IT equipment power.

viii

•“propellant” refers to the chemical substance or combination of substances consumed by a rocket engine to

produce thrust by generating high-velocity exhaust gases.

•“propulsive landing” refers to the process of landing a rocket or spacecraft using its engines to control descent

and achieve a soft, vertical touchdown.

•“radiative cooling” refers to a cooling method that dissipates heat by radiating it into space, often passively, and

is expected to be used in orbital AI compute infrastructure.

•“Raptor engines” refers to high-performance family of engines developed and produced by the Company, such

as those powering the Super Heavy booster and Starship upper stage, designed for efficiency and reusability.

•“reflight” refers to the reuse of a flight-proven rocket booster or upper stage that has successfully completed a

prior space mission, and has been recovered, refurbished, and certified for subsequent launches.

•“return payload mass” refers to the theoretical payload mass that a particular spacecraft is capable of bringing

back to Earth from a specified orbit under specific conditions, which is derived from advanced computer

simulations and performance modelling that apply to particular mission scenarios and trajectory assumptions.

Actual payload that can be returned for a given mission may be different and will vary depending on numerous

mission parameters and operational factors, including mission-specific trajectory requirements, atmospheric

conditions, vehicle and payload configuration, risk profile, and applicable regulatory or range-safety limitations.

•“rideshare” refers to a type of space mission where multiple satellites or payloads from different customers are

launched together on a single rocket, sharing the cost.

•“satellite-to-mobile” refers to a service that provides global cellular connectivity directly to everyday

smartphones via satellites, supplementing terrestrial networks and eliminating mobile dead zones.

•“Service Line” refers to an individual instance of Starlink broadband internet service provisioned under a

subscription plan, generally associated with a specific Starlink User Terminal or group of terminals, and billed

according to Starlink’s service plans and terms of service. The number of Service Lines is distinct from the

number of unique devices, account holders, end users, or physical persons.

•“space economy” refers to economic activities related to the development, production, and operation of goods

and services that utilize or support space-based infrastructure and capabilities, including launch services,

satellite systems, and space-enabled technologies.

•“Space segment” refers to our Space segment, which includes our customer launch operations and offerings

such as Falcon, Dragon, and Starship.

•“SpaceX Bridge Loan” refers to the Bridge Loan Credit Agreement, dated as of March 2, 2026, by and among

the Company, as borrower, the guarantors from time to time party thereto, the lenders from time to time party

thereto and Goldman Sachs Bank USA, as administrative agent and a lender.

•“SpaceX Credit Facility” refers to our Credit Agreement, dated as of February 7, 2025, by and among the

Company, as borrower, the guarantors from time to time party thereto, the lenders from time to time party

thereto and Bank of America, N.A., as administrative agent, as amended by the First Amendment to Credit

Agreement and Waiver, dated as of March 2, 2026, by and among the Company, the lenders party thereto, and

the other L/C Issuers party thereto. In May 2026, the SpaceX Credit Facility was amended to increase the

borrowing capacity and extends the maturity date.

•“spectrum” refers to the range of electromagnetic frequencies used for wireless communication, with licensed

spectrum granting use for specific services.

•“Starlink” refers to our global Low-Earth Orbit satellite constellation and broadband network designed to

deliver high-speed, low-latency internet connectivity worldwide.

ix

•“Starlink Consumer Broadband” refers to a category of Starlink active users encompassing both individual

residential users (households and personal use) and small-to-medium-sized businesses.

•“Starlink Fixed Site” refers to a category of Starlink active users encompassing exclusively enterprise

businesses.

•“Starlink Kit” refers to a set of products needed to connect to the Starlink network, typically including a Starlink

User Terminal and accessories.

•“Starlink Mobile” refers to a service that provides cellular connectivity directly to everyday smartphones via

satellites, supplementing terrestrial networks and substantially reducing mobile dead zones.

•“Starlink Subscriber” refers to a unique Service Line that is directly assigned to a Starlink.com account

registered to a person or entity that does not have a direct, negotiated agreement with the Starlink sales team.

•“Starlink User Terminal” refers to a device developed by the Company that connects to the Starlink satellite

constellation to deliver high-speed, low-latency internet.

•“Starshield” refers to a secure satellite network designed specifically for government customers and national

security applications.

•“Starship” refers to a fully reusable, super heavy-lift launch vehicle. Starship can be used to describe the stacked

vehicle (booster and upper stage) or upper stage only. We expect Starship to commence payload delivery to

orbit in the second half of 2026.

•“Sun-synchronous orbit” refers to a type of polar orbit around a planet in which a satellite passes over any given

point of the planet’s surface at the same local mean solar time, allowing for consistent solar energy capture.

•“Super Heavy” refers to the reusable first-stage booster for the Starship launch vehicle, powered by 33 Raptor

engines.

•“SuperGrok” refers to our subscription-based Grok service that provides users with expanded access to Grok

models and related tools.

•“SuperGrok Heavy” refers to our subscription-based Grok service tier that provides users with expanded access

to Grok models and related tools, including higher usage limits relative to SuperGrok.

•“SuperGrok Lite” refers to our subscription-based Grok service tier that provides users with basic access to

Grok models and related tools.

•“supported accounts” refers to, when used in the context of our X platform and Grok, a human, bot or similar

account that logged into the X platform or Grok. The total number of supported accounts may include fake,

spam or bot accounts if they are active.

•“Tbps” refers to terabits per second.

•“Terafab” refers to a chip manufacturing initiative with a long-term goal of producing one terawatt of compute

hardware each year.

•“terawatt” refers to one trillion watts.

•“terawatt-scale” refers to infrastructure, systems, or facilities that are designed to generate, transmit, or consume

approximately one terawatt or more of electrical power capacity.

•“terrestrial AI compute” refers to artificial intelligence computing infrastructure located on Earth, such as data

centers and supercomputers, used for training and running AI models.

x

•“throughput” refers to the rate at which data or material can be processed or transferred, often referring to

network capacity or production output.

•“tokens” refers to the basic units of text or images processed and generated by an AI model, used to measure AI

workload, throughput, and computational output.

•“watt” is the International System of Units (SI) unit for measuring power, representing the rate of which energy

is transferred, used or generated.

•“X” refers to our real-time information, entertainment, and free speech platform that serves as a foundational

distribution and data engine for the AI ecosystem.

•“xAI” refers to X.AI Holdings LLC or, prior to the xAI Merger, X.AI Holdings Corp., together with its

subsidiaries, as applicable.

•“xAI Gov” refers to our offering that provides government customers with access to Grok models and related

tools for use in governmental applications, workflows, and services.

•“X Premium+” refers to our highest subscription tier for X.

Our Satellite Names

We use a “V” naming convention for our Starlink satellites (such as V1, V2 Mini, and V3). Although we use a

similar “V” naming convention for both our broadband and mobile satellite constellations, these are distinct systems.

Our broadband satellites are designed to deliver high-speed internet services to homes, businesses, and vehicles,

while our mobile satellites are designed to connect directly to cell phones from space. These constellations have

different performance requirements and technical specifications. Please see below the terms used for our satellites

throughout this prospectus:

•“V1 Mobile satellites” refers to our mobile satellites that provide light data, text messaging (SMS), and over-

the-top voice services (e.g., WhatsApp and FaceTime) to mobile devices. V1 Mobile satellites are currently in

orbit and are launched on our Falcon rockets.

•“V2 Mini satellites” refers to our current broadband satellites that provide high-speed internet to homes,

businesses, and vehicles. V2 Mini satellites are currently in orbit and are launched on our Falcon rockets.

•“V2 Mobile satellites” refers to our next-generation mobile satellites, which are designed to provide more

comprehensive satellite-to-mobile services, including broadband data and IoT connectivity and which we expect

to begin deploying on Starship in 2027.

•“V3 satellites” refers to our next-generation Starlink broadband satellites, which are designed to offer one Tbps

of downlink capacity per satellite and which we expect to begin deploying on Starship in the second half of

2026.

1

PROSPECTUS SUMMARY

This summary highlights information contained elsewhere in this prospectus. This summary is not complete and

does not contain all of the information you should consider before investing in our Class A common stock. You

should read this entire prospectus carefully before making an investment decision. You should carefully consider,

among other things, the sections titled “Risk Factors,” “Management’s Discussion and Analysis of Financial

Condition and Results of Operations,” and our consolidated financial statements and the related notes included

elsewhere in this prospectus. Some of the statements in this summary constitute forward-looking statements. Please

carefully consider “Cautionary Statement Regarding Forward-Looking Statements.”

“You want to wake up in the morning and think the future is going to be great—and that’s what being a space-faring

civilization is all about. It’s about believing in the future and thinking that the future will be better than the past. And

I can’t think of anything more exciting than going out there and being among the stars.”

—Elon Musk

Our Mission

Our mission is to build the systems and technologies necessary to make life multiplanetary, to understand the true

nature of the universe, and to extend the light of consciousness to the stars. To do this, we have formed the most

ambitious, vertically integrated innovation engine on (and off) Earth with unmatched capabilities to rapidly

manufacture and launch space-based communications that connect the world, to harness the Sun to power a truth-

seeking artificial intelligence that advances scientific discovery, and ultimately to build a base on the Moon and

cities on other planets.

Overview

Founded in 2002, SpaceX is the only company building the integrated hardware and software infrastructure of the

future across space, connectivity, and AI. At our core, we are builders. We design, manufacture, launch, and operate

products and services built on cutting-edge technologies, including the world’s most advanced rockets and

spacecraft. We safely and reliably transport astronauts, satellites, and other payloads on missions that benefit life on

Earth. Since 2023, we have launched more than 80% of mass to orbit for the world each year with an over 99%

mission success rate with Falcon rockets. We also operate a high-speed, low-latency global broadband data and

communications network powered by approximately 9,600 Starlink broadband and mobile satellites in Low-Earth

Orbit, delivering connectivity to millions of consumer, enterprise, and government customers across 164 countries,

territories, and other markets, as of March 31, 2026. Using our dedicated satellite-to-mobile constellation, we offer

connectivity services, supplementing terrestrial networks and substantially reducing mobile “dead zones” across

approximately 30 countries.

With the potential to improve both space exploration and life on Earth, AI accelerates SpaceX’s mission to make life

multiplanetary, to understand the true nature of the universe, and to extend the light of consciousness to the stars.

xAI, which was founded in 2023 and acquired by SpaceX in early 2026, is now an integral pillar of our vertically

integrated company. We are rapidly constructing AI compute infrastructure—starting on Earth with the goal of

extending to space—at industry-leading pace and cost efficiency. Our infrastructure supports training and inference

for Grok, which has emerged as one of the world’s most advanced frontier models. Grok is designed as a truth-

seeking AI model, built on our founder Elon Musk’s mission to enable humanity to understand the universe. We

believe that accomplishing this mission requires a truth-seeking approach to AI. We define truth seeking as the

active, relentless pursuit of what is objectively true about reality, and grounded in evidence, logic, empirical data,

and first principles thinking. Our goal is to understand and explain what the universe appears to be doing, as

accurately as current knowledge allows. Within two years of its initial model release, Grok achieved frontier-level

performance in scientific reasoning, as measured by its GPQA Diamond score, an industry benchmark that evaluates

AI models on a standardized set of questions written and validated by experts, on a faster timeline than reported by

other leading model providers. Grok also benefits from integration with X, our real-time information, entertainment,

and free speech platform, which serves as a foundational distribution and data engine for our AI ecosystem and

further enhances Grok’s truth-seeking objective.

2

We believe that space represents the largest economic frontier in human history. Connectivity infrastructure in space

is designed to help everyone on Earth have access to education, healthcare, entertainment, and communications, and

to enable people to overcome many traditional limits, such as physical and political borders. We believe AI

infrastructure in space can utilize the virtually limitless power of the Sun and thereby enable the use of AI as a

transformative force for understanding the universe and improving the daily lives of all humans. We believe the

convergence of these areas will enable an unprecedented expansion in the global economy, leading to an age of

abundance. Our innovations and technological advancements are redefining industries on Earth, while we aim to

create new ones on the Moon, Mars, and beyond. We are truly building the infrastructure of the future.

•Space. SpaceX is the only company that has cracked the code on accessing space at scale, revolutionizing an

industry characterized by decades of stagnation, risk aversion, and economically perverse cost structures.

SpaceX upended this paradigm through the application of first-principles thinking, which rejects industry

assumptions and builds solutions based on the fundamental laws of physics. Our intense, mission-driven,

engineering-first culture and focus on extreme vertical integration have propelled us to achieve what many

deemed impossible. We pioneered high-cadence, reliable, and affordable access to space with our Falcon family

of rockets. In 2015, we established at least a 10-year lead over the industry by successfully landing our first

Falcon 9 booster back from space before anyone else. Space flight that historically cost billions per launch now

costs in the tens of millions, fundamentally reducing the cost of space access and providing the opportunity to

build new enterprises in space.

•Connectivity. Since activating service for customers in 2020, Starlink has rapidly expanded global access to

high-speed internet, prioritizing underserved rural and remote communities worldwide. While building

terrestrial networks in such communities can be prohibitively expensive, Starlink is capable of delivering

broadband connectivity anywhere on Earth with just a Starlink Kit. As of March 31, 2026, we had

approximately 9,600 Starlink broadband and mobile satellites in Low-Earth Orbit, operating the world’s most

advanced broadband constellation providing internet connectivity to approximately 10.3 million Starlink

Subscribers across 164 countries, territories, and other markets. In January 2024, we also began deploying our

Starlink Mobile constellation that utilizes separate Starlink satellites with satellite-to-mobile capabilities,

substantially reducing mobile “dead zones” around the world. As of March 31, 2026, our dedicated satellite-to-

mobile constellation of approximately 650 V1 Mobile satellites provides satellite-to-mobile data, over-the-top

voice, and messaging services to approximately 7.4 million monthly unique devices across approximately 30

countries.

•AI. We were the first company to deploy a coherent gigawatt-scale AI training cluster. For complex reasoning

and agentic workloads, compute is directly correlated with the quality of intelligence and task completion speed.

In under two years, we have established a dual advantage in both cost efficiency and deployment speed at scale.

By owning the compute infrastructure and vertically integrating across the full AI stack, we can train and iterate

our frontier models at lower cost and higher velocity and accelerate development cycles. This eliminates

external bottlenecks and drives rapid, continuous improvements in model performance. We believe this

combination of our state-of-the-art AI compute infrastructure, our truth-seeking frontier model, and our access

to real-time data on X creates a significant strategic advantage. Our integrated AI platforms across Grok and X

have over 1.3 billion supported accounts active in the last twelve months ended March 31, 2026, including

approximately 550 million MAUs and generating approximately 350 million daily posts. Of our MAUs, we had

approximately 117 million MAUs that used Grok’s AI features as of March 31, 2026. Grok’s deep integration

with X enables freshness, relevance, and contextual awareness that we believe is a competitive differentiator.

This direct, real-time access to the information and human discourse on X enhances Grok’s truth-seeking

capabilities by grounding outputs in up-to-date knowledge and diverse viewpoints. As a result, we believe Grok

can deliver the most objective and relevant insights and best serve high-frequency, high-value use cases across

consumer and enterprise AI applications.

We have created distinct new markets across the space, connectivity, and AI industries by building the integrated

hardware and software infrastructure of the future and by combining our broad range of capabilities. For example,

SpaceX’s recent acquisition of xAI unites SpaceX’s launch capabilities and global connectivity network with xAI’s

AI development capabilities. Specifically, we believe SpaceX’s reusable rockets, scaled satellite manufacturing, and

operational expertise can enable the cost-effective and rapid deployment of massive AI compute satellite

3

constellations—with potentially millions of satellites—for orbital data centers. We believe these AI compute

satellites in Sun-synchronous orbit will be able to handle energy-intensive AI workloads, such as inference demand,

at far greater scale and efficiency than terrestrial alternatives, with Starlink providing low-latency, global

connectivity linking these orbital AI systems to people around the world and delivering real-time intelligence. We

expect to begin deploying our orbital AI compute satellites as early as 2028.

Our financial results reflect the strength of our operating model and our ability to create and scale multiple new

businesses:

•For the three months ended March 31, 2026, we generated revenue on a consolidated basis of $4,694 million,

loss from operations of $(1,943) million and Adjusted EBITDA of $1,127 million. In 2025, we generated

revenue on a consolidated basis of $18,674 million, loss from operations of $(2,589) million and Adjusted

EBITDA of $6,584 million. Our Space and Connectivity segments contributed the substantial majority of our

consolidated revenue in the three months ended March 31, 2026 and the year ended December 31, 2025,

demonstrating the benefits of their scale and operating leverage in our vertically integrated business model;

•For the three months ended March 31, 2026, our Space segment generated revenue of $619 million, loss from

operations of $(662) million, and Segment Adjusted EBITDA of $(351) million. In 2025, our Space segment

generated revenue of $4,086 million, loss from operations of $(657) million, and Segment Adjusted EBITDA of

$653 million. Additionally, our Space segment funded $930 million and $3,004 million in research and

development expense during the three months ended March 31, 2026 and the year ended December 31, 2025,

respectively, for our next-generation Starship launch vehicle program. Starship is designed to enable a step-

function change in our launch capability across reusability, payload capacity, and launch cadence, and is the key

enabler of our long-term growth strategy by unlocking entirely new categories of missions;

•For the three months ended March 31, 2026, our Connectivity segment generated revenue of $3,257 million,

income from operations of $1,188 million, and Segment Adjusted EBITDA of $2,087 million. Our Connectivity

segment, primarily driven by Starlink, generated revenue of $11,387 million, income from operations of $4,423

million, and Segment Adjusted EBITDA of $7,168 million in 2025, representing year-over-year growth of

49.8%, 120.4%, and 86.2%, respectively, benefiting from subscriber growth, increasing enterprise adoption, and

continued improvement in network efficiency;

•In our newly acquired AI segment, we plan to prioritize growth and investment to capture significant

opportunities in AI applications and compute infrastructure. For the three months ended March 31, 2026, our AI

segment generated revenue of $818 million, loss from operations of $(2,469) million, and Segment Adjusted

EBITDA of $(609) million. In 2025, our AI segment generated revenue of $3,201 million, loss from operations

of $(6,355) million, and Segment Adjusted EBITDA of $(1,237) million, reflecting its earlier stage of

development and continued investments to support long-term growth opportunities in AI; and

•For the three months ended March 31, 2026, capital expenditures for our Space segment was $1,052 million, for

our Connectivity segment was $1,332 million and for our AI segment was $7,723 million. In 2025, capital

expenditures for our Space segment was $3,832 million, for our Connectivity segment was $4,178 million and

for our AI segment was $12,727 million. 

Segment Adjusted EBITDA is a non-GAAP measure. Please refer to the section titled “Management’s Discussion

and Analysis of Financial Condition and Results of Operations—Non-GAAP Financial Measures” for additional

information on our non-GAAP financial measures, including reconciliations of Segment Adjusted EBITDA to

segment income (loss) from operations, the most directly comparable GAAP measure.

Why This Matters Now

For the entirety of its existence, human civilization has lived on a single celestial body: Earth. The current paradigm,

in which human civilization is confined to one planet, exposes humanity to existential threats that are unpredictable

and uncontrollable on a planetary scale. By moving beyond the only home we have ever known, we ensure species-

level redundancy and that the light of consciousness will not be tied to a single planet subject to the inevitable

hazards of a harsh and vast universe. We do not want humans to have the same fate as dinosaurs. We want to give

4

them a reason to look ahead with excitement, with the prospect that we are entering an age of abundance with an

endlessly prosperous and exciting future.

For decades, a reality where humanity travels between the planets and the stars has felt tantalizingly close but still

locked in the pages and screens of science fiction. We are capable of better understanding the universe, exploring the

universe, and ultimately making life multiplanetary across the universe. We are becoming a civilization with the

ability to reach beyond Earth’s cradle and begin to inhabit other worlds. While we remain dedicated to this

fundamental mission, our progress in accessing space continues to yield opportunities that enrich life on Earth. For

example, by dramatically reducing the cost of access to space, we have been able to expand our mission to address

some of the Earth’s most pressing challenges, including bridging the digital divide by aiming to connect over three

billion unconnected people to the internet and humanity’s collective knowledge.

The rapid emergence of the AI era intensifies the urgency of our mission, as AI has the potential to accelerate not

only space exploration, but also transformative societal advancements on Earth. However, AI’s ability to

revolutionize human potential is directly dependent on meeting exponentially increasing resource demands. On

Earth, the massive expansion of data center capacity to support growing compute demand is significantly outpacing

electricity generation, which was effectively flat in the United States for approximately 15 years, growing at a

compound annual growth rate of 0.1% from 2008 to 2023. Despite the recent increase in electricity demand from AI

data centers, electricity generation in the United States has grown at an annual rate of less than 3% between 2023

and 2025, while electricity generation in China has grown at approximately twice that rate in the same time period.

This supply and demand imbalance is already imposing unsustainable strains on terrestrial power grids, supply

chains, and the environment. The Sun contains approximately 99.8% of the solar system’s energy and, as a result,

we believe it is the only truly scalable solution to terrestrial energy constraints in the age of AI. Harnessing this

energy in space is considerably more efficient than on land. Space-based solar arrays can generate more than five

times the energy per unit area of terrestrial solar due to continuous illumination, lack of atmospheric interference,

and optimal orientation. SpaceX is well-positioned to capture this space-based solar energy through our ability to

rapidly access Sun-synchronous orbit through our satellite manufacturing scale and launch capability. As a result,

we are expanding our footprint and harnessing the vast resources of space that are essential to sustaining

technological development. Our goal is to ensure that AI becomes a force for human flourishing and a benefit to

civilization, rather than a catalyst for terrestrial resource depletion and instability.

We believe that our current space efforts will catalyze transformative breakthroughs that could reshape terrestrial

industries and lead to the emergence of new trillion-dollar markets on the Moon, Mars, and beyond. In particular, we

believe our goal of establishing a lunar presence will enable terawatt-scale annual AI compute growth, support

deeper space exploration and industrialization, and serve as a stepping stone to establishing a civilization on Mars.

We believe the next paradigm shift for humanity is the creation of a resilient, perpetually expanding spacefaring

civilization that drives continuous innovation across new frontiers, ultimately propelling us to Kardashev Type II

status—we believe we are capable of unlocking an era of unprecedented economic expansion, while also

contributing to the safeguards of humanity’s future against existential risk.

Who We Are

SpaceX combines the most transformative and critical technologies in human history, including reusable rockets, a

fully global internet service, satellite-to-mobile communications, a real-time information, entertainment and free

speech platform, and a truth-seeking AI system designed to accelerate scientific discovery and augment human

capabilities.

Our Unparalleled Launch Capabilities

Since our founding in 2002, SpaceX has cracked the code on accessing space at scale, transforming an industry

characterized by decades of stagnation, risk aversion, and economically perverse cost structures. We design,

manufacture, launch, and refurbish reusable launch vehicles that provide cost-efficient, reliable, and high-cadence

access to space for our own purposes as well as for third-party commercial and government customers. Our

extensive vertical integration and end-to-end control over the entire value chain, from design to launch to operations,

allows us to achieve unprecedented speed and cost efficiency.

5

As of March 31, 2026, SpaceX had launched a total mass to orbit of approximately 7,400 metric tons with an over

99% mission success rate across our Falcon rockets. We have completed approximately 650 orbital space launches,

and over 540 of those launches were completed by a flight-proven Falcon rocket. With the first successful launch of

Falcon 1 in 2008, we became the first private company to successfully launch a liquid-fueled rocket to Earth’s orbit.

In December 2015, we achieved what many deemed impossible: landing a rocket launched to space back on Earth.

By 2017, we were routinely recovering and reusing the Falcon 9 first-stage booster post-launch, delivering another

step-function drop in space access costs via groundbreaking reusability. As of March 31, 2026, our Falcon 9 rockets

have demonstrated the ability to refly a first-stage 34 times. With the future deployment of Starship, which is

designed to be the world’s first fully and rapidly reusable spacecraft, we aim to reduce the cost to reach orbit by 99%

or more relative to the historical average launch cost, establishing the most affordable and scalable path to creating

new opportunities in space, such as orbital AI compute and Mars exploration.

Our principal launch vehicles and spacecraft include:

•Falcon 9. As the world’s first orbital-class rapidly reusable rocket, Falcon 9 was first launched in 2010 and has

a payload capacity to LEO of approximately 23 metric tons when fully expendable. Falcon 9 has completed

approximately 620 orbital space launches as of March 31, 2026, and an over 99% mission success rate.

According to NASA, the first version of Falcon 9 in 2010 reduced launch cost to approximately $2,700 per

kilogram, approximately 85% less than the historical average launch cost of $18,500 per kilogram.

•Falcon Heavy. Falcon Heavy first launched in 2018 when it put a Tesla all-electric sports car (“Tesla

Roadster”) and its mannequin passenger, known as Starman, into orbit around the Sun. With a payload capacity

to LEO of approximately 64 metric tons, Falcon Heavy is a partially reusable super heavy-lift launch vehicle

designed to deliver large payloads to orbit. Falcon Heavy is one of the most powerful operational rockets in the

world measured by liftoff thrust, with 11 launches as of March 31, 2026 and a 100% mission success rate.

•Dragon. Launched by Falcon 9 in 2012, our Dragon spacecraft became the first commercial spacecraft to

deliver cargo to and from the International Space Station, an orbiting laboratory that serves as a research facility

and destination for human spaceflight, and, eight years later, the first privately built vehicle to fly humans to the

orbiting laboratory. Since 2020, our Dragon spacecraft has safely flown 78 crewmembers from 20 countries.

•Starship. First launched in 2023, Starship is designed to be a fully reusable, super heavy-lift launch vehicle.

Starship V3 is designed to deliver 100 metric tons to Earth’s orbit in a fully reusable configuration while

enabling rapid turnaround times akin to commercial aviation. Future generations of Starship are being designed

to double this payload capacity. To date, we have executed 11 Starship flight tests. We have also scheduled a

12th flight test, which will debut the next generation Starship vehicle and Super Heavy booster, powered by the

next evolution of our Raptor engine and launching from a newly designed pad at Starbase. We expect Starship

to commence payload delivery to orbit in the second half of 2026. We have achieved innovative milestones

such as catching a booster using “chopstick” arms on the same tower it launched from. We expect this

capability will facilitate rapid refurbishment and reuse, allowing for multiple launches per day at reduced costs.

Upon achieving rocket reusability, we recognized the immense potential of our launch business to enable new

revenue streams. This led to the development of Starlink, our global satellite internet constellation, consisting of

thousands of LEO satellites designed to provide high-speed, low-latency broadband connectivity to underserved

areas worldwide. Although the concept of using satellites for global internet connectivity dates back decades,

technical challenges and the prohibitive cost of accessing space and deploying the satellites required for capacity and

global coverage historically rendered attempts to provide such connectivity economically unviable. Within three

years of our first satellite launch in 2019, we solved the technical and production challenges of the satellites, and

within five years, we had deployed the largest LEO constellation in existence. Today, Starlink is the sole low-

latency network available globally. By combining increasing launch cadence, expanding cargo capacity, and

declining unit costs—driven by rapid reusability—we have generated a compounding competitive advantage. This

not only fortifies our core business, but also provides vast new market opportunities uniquely enabled by space.

6

Our Leading Capabilities Across Space, Connectivity, and AI

Space.While our launch capabilities support our other businesses, such as Starlink Consumer Broadband and

Starlink Mobile, we also sell launches to third-party customers. We offer launch services to commercial, civil,

international and government customers through our reusable Falcon 9 and Falcon Heavy rockets for satellite, cargo,

and crew missions. We are the primary launch provider for the U.S. government. In 2025, we launched 11 of 12

National Security Space Launch (“NSSL”) medium and heavy lift missions and all five U.S. crew and cargo

missions to the International Space Station for NASA.

Connectivity. Our Connectivity business includes Starlink Consumer Broadband, Enterprise Solutions, Government

Solutions, and Starlink Mobile.

•Starlink Consumer Broadband. We operate the world’s largest and most advanced space-based internet

broadband service. We provide fiber-like download speeds—at a median of 225 Mbps during peak hours for

residential users as of March 31, 2026—and the technological capability to provide service everywhere on

Earth, including the poles. This service quality is enabled by our vast network of approximately 9,600 Starlink

broadband and mobile satellites in Low-Earth Orbit, which accounted for approximately 75% of all active

maneuverable satellites in orbit as of March 31, 2026. We expect to commence deploying our next-generation

V3 satellites, designed to offer one Tbps of downlink capacity per satellite, using Starship in the second half of

2026. We expect that a single Starship launch will be capable of deploying up to 60 V3 satellites to LEO,

representing a potential twenty-fold increase in Starlink downlink capacity deployed relative to a Falcon 9

launch.

•Enterprise Solutions. SpaceX is a critical partner to a wide array of enterprises. We offer Starlink’s high-

speed, low-latency, reliable internet services to enterprise customers across industries including construction,

agriculture, retail, telecom, hospitality, aviation, maritime, and land mobility. Starlink’s unique capabilities are

well‑suited for deployments across field offices, remote worksites, research stations, drilling rigs, rural

hospitals, aircraft, cruise ships, trains, and hotels. We also serve a broad fixed‑site customer base across

industries such as retail and financial services that require high availability for critical operations as well as

reliable connectivity in remote or hard-to-serve locations.

•Government Solutions. For our government customers, we provide high-speed, resilient connectivity for

public services, social impact, humanitarian efforts, and disaster response in even the most remote and

challenging environments. Separately with Starshield, we have leveraged our commercial LEO satellite

constellation engineering learnings and operational experiences to develop a secure, dedicated satellite network

designed specifically for United States Government customers and national security applications.

•Starlink Mobile. We provide satellite-to-mobile connectivity, supplementing terrestrial networks and

substantially reducing mobile “dead zones” across approximately 30 countries. Through our partnerships with

approximately 30 MNOs on six continents, we enable consumers, businesses, and public-sector customers to

use their existing phones in more places, support critical connectivity during disasters and power outages, and

open new applications for low-bandwidth mobile and IoT devices.

AI. We operate a highly vertically integrated AI platform.

•AI Compute Infrastructure. xAI has established a leading position in building and scaling terrestrial AI

compute infrastructure, becoming the first company to deploy a coherent gigawatt-scale AI training cluster. We

own and operate what we believe to be the largest AI training data center clusters on Earth, including

COLOSSUS and COLOSSUS II. The addition of Terafab, a chip manufacturing initiative with Tesla and Intel,

aims to further extend our vertical integration to chip design and manufacturing to alleviate potential future chip

shortages at SpaceX, optimize compute performance, and potentially reduce overall compute costs. In

connection with such collaboration, we have agreed with Tesla on a general framework for the future

development of Terafab. Any specific projects undertaken pursuant to this framework will be subject to separate

negotiations and agreements (including any development timelines, milestones and capital expenditures) and

have not yet been determined. We believe that the key constraints in the continued growth of AI are physical—

7

chip manufacturing, data center infrastructure, and power generation; the future of AI will be determined by the

control of the physical stack.

•Truth-Seeking Frontier Model. Since launching Grok-1 in November 2023, we have released four major

versions and notable variations thereof, achieving one of the fastest iteration cycles in the industry. Within two

years of its initial model release, Grok achieved frontier-level performance in scientific reasoning, as measured

by its GPQA Diamond score, an industry benchmark that evaluates AI models on a standardized set of

questions written and validated by experts, on a faster timeline than reported by other leading model providers.

Building on this trajectory, we expect to continue scaling Grok through subsequent generations. Ongoing

training of next‑generation models is expected to scale toward multiple trillions of parameters, which could

represent a step change in reasoning in depth and overall intelligence.In this context, the number of parameters

refers to the scale of the model, where parameters are the internal numerical values, such as “weights,” that are

adjusted during training to enable the model to recognize patterns and relationships in data. A larger number of

parameters generally allows the model to capture more complex relationships, store greater amounts of

knowledge, and achieve higher levels of reasoning capability. This accelerated rate of innovation stems from

our highly vertically integrated stack: full ownership of training infrastructure; access to the world’s most

powerful compute clusters; and relentless focus on truth seeking and real-world utility. A key competitive

differentiator is Grok’s deep integration with X, enabling proprietary access to a real-time information stream of

approximately 350 million daily posts, which enhances freshness, relevance, and contextual awareness for

Grok. This direct, real-time access to the information and human discourse on X enhances Grok’s truth-seeking

capabilities by grounding outputs in up-to-date knowledge and diverse viewpoints.

•Consumer and Enterprise Applications. We leverage our leading frontier models and compute infrastructure

to deliver consumer and enterprise applications. Together with Tesla, we are also developing Macrohard, an

agentic AI platform designed to be capable of fully emulating digital workflows and augmenting human

operation of computersusing sophisticated autonomous agents. We believe Macrohard will have the potential to

fundamentally transform how companies are structured and operate, thereby allowing dramatic increases in

human productivity.

Our Repeatable Business Model

Our business model is built on a repeatable, engineering-driven framework that combines our unparalleled launch

capabilities, extreme vertical integration, rapid iteration, and disciplined capital investment to create durable, large-

scale businesses. We execute this framework through the following core principles:

1.Leverage our unparalleled launch capabilities to enable massive scale;

2.Identify and create new trillion-dollar market opportunities;

3.Design a solution with world-class engineering and first-principles thinking;

4.Apply “The Algorithm” (make less dumb, delete, optimize, accelerate, automate);

5.Vertically integrate all the way to the end customer;

6.Continuously drive cost down and throughput up; and

7.Generate significant cash flow and reinvest in the future.

Our Engineering-First Culture

We are able to achieve transformative technological breakthroughs because we accept only the laws of physics as

the limiting factors to our work and mission. Our core approach is deeply rooted in first-principles thinking, which

rejects any preconceived notions or experience-based norms. We have a track record of achieving what many have

deemed impossible. Some of our industry-defining achievements and historic milestones include:

•The first private company to develop and launch a liquid-fuel rocket to reach orbit (2008);

•The first private company to successfully dock a private spacecraft with the International Space Station (2012);

•The first to successfully propulsively land (2015) and refly orbital-class rocket boosters (2017);

•The first to begin deploying a large-scale LEO broadband satellite constellation (2019);

•The first private company to transport astronauts to orbit, returning America’s ability to fly astronauts to and

from the International Space Station (2020);

8

•The first to manufacture consumer-grade phased-array user terminals at scale (2022);

•The first to deploy a large-scale LEO satellite-to-mobile constellation (2025);

•The first to build a gigawatt-scale AI training cluster and largest coherent supercomputer (2026);

•The first gigawatt-scale Megapack battery installation (2026); and

•The only company capable of building orbital AI compute at scale.

Our AI Compute Infrastructure Advantage and Growth Strategy

Why Compute Matters. We believe AI leadership will be defined by the ability to rapidly scale compute capacity to

support exponential usage growth and frontier intelligence. The training and inference demanded by advanced AI

models require substantial computational resources. Reasoning models introduced in 2024 demonstrated that

allocating more computational resources and giving models more time to process during inference directly leads to

higher-quality intelligence. In addition, compute infrastructure with end-to-end, cluster-level coherence through tight

integration across software and hardware systems enables more efficient, stable, and higher-fidelity training and

inference at scale—ultimately enhancing model intelligence and performance. Within inference, we expect

computationally-intensive reasoning, agentic, and multi-modal workloads will continue to grow as a portion of

overall usage. We therefore believe operators with superior model-to-compute integration—the ability to efficiently

support and allocate compute across both training and inference workloads—are best positioned to win the AI race.

Self-Reinforcing Network Effects Among Lower Cost Per Token, Model Quality, and User Adoption.AI systems

are ultimately constrained or differentiated by the cost, speed, and scale at which they can generate and process

tokens. A “token” represents the fundamental unit of data consumed and produced by modern AI models. This is

because lower cost per token enables more frequent model training, larger and more sophisticated models, longer

chains of processing for reasoning and agentic workloads, and significantly higher inference volumes at

economically viable prices. This dynamic directly impacts model quality, responsiveness, and accessibility, while

also determining the ability to serve the rising global demand across consumer, enterprise, and mission-critical AI

applications. This creates a self-reinforcing advantage in which lower token costs drive greater model quality and

user adoption, reinforcing AI leadership.

Cost of Compute is the Main Driver of Cost Per Token. The total cost per token is determined by the efficiency,

availability, and unit economics of the underlying compute and the cost of building and operating compute

infrastructure. Improvement in the cost of building and operating this compute infrastructure—whether through

lower data center construction cost, lower power infrastructure cost, shorter time to grid interconnection, or higher

cluster-level throughput—translates directly into lower cost per token. Accordingly, for a given level of intelligence,

we expect the long-term economics of AI companies to be driven by the ability to consistently deliver bleeding-edge

compute at the lowest possible cost per token. Put simply, we view cost per token as a function of three primary

inputs—the underlying AI model, the compute hardware, and energy, and we expect to have a competitive

advantage in the latter two cost components. We believe we have a pathway over time that will significantly reduce

compute hardware costs through continued vertical integration and development of proprietary chips, building on

our experience designing custom silicon for our Starlink satellites. We also expect that the marginal cost of energy

for our AI compute satellites will be minimal because our satellites are powered by solar arrays in space. By driving

the energy component to minimal levels and pursuing improvements in compute hardware cost, we believe we can

achieve a meaningfully lower overall cost per token in the future.

We Have a Dual Speed and Cost Advantage in Terrestrial AI Compute. We own and operate what we believe to be

the largest AI training data center clusters on Earth. Our AI compute facilities, COLOSSUS and COLOSSUS II,

collectively provide approximately 1.0 gigawatt of compute power, with additional power capacity available for data

center operations. Our first-principles thinking enables us to build coherent compute at scale and at rapid speed with

lower costs than most other companies in the industry. In order to bring compute clusters online as fast as possible,

we employ a vertically integrated, nimble approach to construction. We brought the first cluster of COLOSSUS

online in 122 days, repurposing the shell of an existing factory, and the first cluster of COLOSSUS II online even

faster in 91 days. As an illustrative comparison,  an industry benchmark to bring online a 100 megawatt greenfield

data center is approximately two years. We also demonstrated a significant improvement in cost efficiency,

achieving data center construction costs for COLOSSUS II that are considerably lower than industry benchmarks on

a per megawatt basis.

9

We Believe Orbital AI Can Accelerate Time to Power and Reduce Token Costs.The Sun contains approximately

99.8% of the solar system’s energy and offers what we believe is the only truly scalable solution to the challenge of

accelerating demand for compute relative to terrestrial energy constraints. The logical path forward is to move

power-intensive AI workloads into orbit, where solar energy is near-constant and uninterrupted. With such

accessibility to energy, we believe that our launch business will enable us to consistently activate the highest

performing hardware before our competitors without such access, shrinking the timeline to useful tokens on

bleeding-edge hardware and sustaining our token cost advantage. We believe SpaceX is uniquely positioned to

deploy and operate data centers in orbit that can eventually achieve a lower cost than terrestrial data centers over

time due to our extreme vertically integrated approach across launch, satellite manufacturing at scale,network

connectivity, and terrestrial data center expertise.

We Believe We Are Well-Positioned to Deliver Orbital AI Compute.We believe orbital AI compute is an incredibly

difficult technical challenge that only we can solve at scale in the near term. We are the only company that has

already accomplished the key technical challenges associated with evolving connectivity satellites into AI compute

satellites. In our view, we are well-positioned to deliver a full-scale AI compute satellite constellation. Significant

work remains, but we are confident in our singular leadership position.

•We have unmatched satellite launch capabilities to enable deployment at scale. Deployment of 100

gigawatts per year via satellites carrying over 100 kilowatts of compute power per metric ton will require

thousands of launches per year and the transport of approximately one million metric tons to orbit annually. The

fully reusable nature of Starship positions us to be capable of launching this level of mass. Starlink Broadband

V1 and V2 Mini satellites have already demonstrated launch survivability and high reliability under vibration,

shock, g-loads, acoustic stress, and vacuum exposure, achieving 99.9% average uptime.

•We have already solved many of the significant technical hurdles to evolving connectivity satellites into

AI compute satellites. Through our leading expertise of connectivity satellites—including mass production,

deployment, network operations, and inter-satellite lasers and mesh connectivity—we have already solved the

hardest part in the development of AI compute satellites. Because AI compute satellites represent an evolution

of spacecraft engineering already demonstrated through Starlink, we believe development of AI compute

satellites will be easier for us than for anyone else. Our existing Starlink constellation is another crucial enabler

of orbital AI compute, as its global network allows data from our AI compute satellites to reach ground stations

anywhere on Earth.

•We will use our proven Starlink in-orbit technology to optimize our orbital AI compute. In order to

operate orbital AI compute satellites, we plan to build on our vast experience of operating approximately 9,600

Starlink broadband and mobile satellites in Low-Earth Orbit. In 2025 alone, Starlink satellites proactively

performed over 1,000 automated collision avoidance maneuvers per day guided by this technology to safely and

efficiently operate the constellation. This operating model gives us control over workload placement across

Earth and space while maintaining resilience through redundancy and fail safe systems. A high degree of

controllability will allow the satellite to be optimized for brightness mitigation, disposal, and other modes of

operation.

•We can manufacture our AI compute constellations at scale with rapid upgrade cycles. We have built one

of the largest satellite manufacturing operations in the world. Our vertically integrated approach with limited

reliance on third-party suppliers will be key to our mass-scaling efforts and should allow us to deploy the latest

AI processors. We believe SpaceX will be the first and only company to manufacture satellites at the scale of

automotive manufacturing.

•We are building chip manufacturing capabilities to scale our access to AI compute hardware. We

announced a collaboration with Tesla in March 2026 to build the Terafab initiative with a long-term goal of

producing one terawatt of compute hardware each year. In connection with such collaboration, we have agreed

with Tesla on a general framework for the future development of Terafab. Intel joined the project in April 2026

and is expected to contribute its expertise in designing, fabricating, and packaging ultra-high performance chips

to help Terafab scale. Any specific projects undertaken pursuant to this framework will be subject to separate

negotiations and agreements (including any development timelines, milestones and capital expenditures) and

10

have not yet been determined. With this internal manufacturing capability, we plan to alleviate potential future

chip shortages at SpaceX, especially as we develop orbital AI at scale, and design chips that are optimized for

the space environment.

•We can leverage our terrestrial experience to build and operate compute clusters and AI workloads at

scale. We believe our experience operating compute infrastructure on Earth provides the technical and

operational foundation to extend these capabilities into orbit. For example, we plan to subject compute hardware

to extensive pre-deployment testing on Earth to identify early life failures before launch to reduce in-orbit

disruption. For compute hardware that does fail, we plan to leverage existing Starlink fleet management

software to reallocate traffic to other satellites and prevent cluster-level downtime.

We Believe Our Infrastructure is a Distinct Advantage in Delivering Superior AI.We expect the combination of

competitive cost per token, our ability to deploy and operate data centers in orbit, and our strength in connectivity to

result in more scalable intelligence that is accessible globally at high speeds.

Our Strengths

•Global Leadership in Orbital Launch Services

•Unrivaled Satellite and Connectivity Platform across Design, Manufacturing, Deployment, and Operations

•Truth-Seeking AI Model Enhanced by Real-Time Data

•Extreme Vertical Integration Enabling High Velocity and Superior Cost Efficiency at Scale

•Unique Ability to Scale New Trillion-Dollar Markets Across Space, Connectivity, and AI

•Business Models that Are Incredibly Difficult to Replicate

•Mission-Driven Culture and World-Class Talent

Our Growth Strategies

Space

•Increase launch payload capacity

•Establish the lunar economy, including cargo transport, manufacturing, and energy production on the Moon

Connectivity

•Grow Starlink Broadband customers

•Expand our Starlink Mobile offering

•Increase the capacity of our constellations

AI

•Grow consumer AI platform monetization

•Grow X monetization

•Deepen enterprise and government adoption

•Increase the scale of our terrestrial power and AI compute infrastructure

•Deploy orbital AI compute at scale

•Design and manufacture our own chips

•Launch digital human augmentation

Future Markets

•Point-to-point terrestrial travel

•Space tourism

•In-orbit manufacturing

•Passenger and cargo transport to the Moon and Mars

•Energy production on the Moon and Mars

•Manufacturing capabilities on the Moon and Mars

•Asteroid mining

11

Our Market Opportunity

We believe we have identified the largest actionable total addressable market (“TAM”) in human history. We

estimate that our quantifiable TAM is $28.5 trillion, consisting of $370 billion in Space from space-enabled

solutions; $1.6 trillion in Connectivity across $870 billion in Starlink Broadband and $740 billion in Starlink Mobile

as well as additional opportunities in enterprise and government; $26.5 trillion in AI across $2.4 trillion in AI

infrastructure, $760 billion in consumer subscriptions, $600 billion in digital advertising, and $22.7 trillion in

enterprise applications. For illustrative purposes of sizing our addressable market opportunity, we exclude China and

Russia from our global estimates.

SpaceX’s Estimated TAM by Segment

Our Challenges

We face a number of challenges relating to our business and growth strategy and, ultimately, the achievement of our

mission to make life multiplanetary, understand the true nature of the universe, and extend the light of consciousness

to the stars. The pursuit of our mission drives our decision-making and forms the foundation of our business plan,

which is  predicated on building, commercializing, and operating services and products at a scale that has not

previously been achieved. This objective requires us to develop and integrate complex and novel technologies,

develop new processes and infrastructure, and coordinate across multiple suppliers, contractors, regulators, and

stakeholders. Because we are attempting to execute at a scale for which there is no precedent, we face heightened

uncertainty with respect to design, engineering, procurement, construction, commissioning, and operational

performance. In particular, our ability to execute our growth strategy is highly dependent on the successful

development and scaling of Starship and the ability to increase our launch cadence, both of which are subject to

challenges and uncertainties inherent in the development and deployment of new and complex technologies.

Additionally, many of our initiatives described above under “Our Growth Strategies,” including those to develop

orbital AI compute at scale, manufacture AI chips at scale, establish a lunar economy, transport humans and cargo to

the Moon and Mars, and develop human augmentation systems, involve significant technical complexity, unproven

technologies or technologies that do not exist, and such initiatives may not achieve commercial viability. Many of

the innovative products and services described elsewhere in this prospectus may ultimately be unsuccessful and may

require great expense, innovations not yet achieved or technologies not yet developed. As a result, the timeline for

certain of our initiatives involving unproven or new innovations, including our goal of deploying 100 gigawatts of

12

annual compute power to orbit, the establishment of a lunar economy and interplanetary industrialization, and the

launch cadence required to achieve these goals may be difficult or impossible to determine. Our growth strategy may

take longer to execute than anticipated, and you may not realize a return on your investment within the timeframe

you anticipate, or at all.

In addition, a portion of our anticipated market opportunities is associated with industries described above under

“Future Markets.” Certain of these industries, such as space tourism and cargo transport to the Moon, are still

emerging. Others, including in-orbit manufacturing, passenger transport to the Moon, passenger and cargo transport

to Mars, energy production on the Moon and Mars, manufacturing capabilities on the Moon and Mars, and asteroid

mining, do not exist today. While we believe these industries will develop over time, the manner in which they

emerge, including the timing of commercialization, the scale and pace of adoption, and the applicable competitive,

technical, regulatory, geopolitical, and economic frameworks may differ materially from our current expectations.

Our Space, Connectivity, and AI segments are also subject to the following challenges and uncertainties, among

others.

•Space: Our growth strategy depends on our ability to increase our launch cadence and payload capacity, which

is dependent on the successful development of Starship at scale. Unexpected design modifications, supply chain

disruptions, anomalies, environmental issues, and other unforeseen technical challenges could result in delays or

failures to deploy Starship on our anticipated schedule, which would delay or impede our ability to achieve our

other business objectives, such as the deployment of our next-generation satellites, the expansion of our

satellite-to-mobile connectivity services, and deployment of in-orbit AI compute infrastructure.

•Connectivity: Our satellite connectivity, including our global satellite-to-mobile connectivity services under

Starlink Mobile, depend on access to radio frequency spectrum and authorizations from the Federal

Communications Commission (the “FCC”) in the United States and telecommunications regulators in other

countries. Acquiring the necessary authorizations can be a complex and time-consuming process. Without these

licenses and approvals, we cannot generally offer connectivity services in a given market. Spectrum access itself

is limited and highly regulated. Additionally, the growth of our connectivity services depends on our ability to

increase market awareness and acceptance of connectivity through Starlink across numerous international

markets, each with its unique challenges.

•AI: Our AI business is in a relatively early stage, it is being integrated into our organization, its business

strategy is still developing, and it will require significant capital expenditures to fund compute, infrastructure

and power generation, model training, and product development. Additionally, our AI business is subject to

challenges inherent in a nascent, highly competitive, capital intensive and rapidly changing industry. These

include the potential for disruptive technological change, evolving industry and regulatory standards, the

emergence of new and well-funded competitors, frequent new product and service introductions, and changing

customer demands.

Any number of these challenges, and others that may be currently unknown to us, could have a negative impact on

our business, financial condition, and results of operations. For a discussion of the challenges, risks, and limitations

that could harm our future prospects, please refer to “Cautionary Note Regarding Forward-Looking Statements,”

“Risk Factors,” and “Management’s Discussion and Analysis of Financial Condition and Results of Operations”

included elsewhere in this prospectus.

Recent Developments

Collaboration with Cursor

In April 2026, we entered into a compute and option agreement with Anysphere, Inc., doing business as Cursor, a

San Francisco-based private software company (“Cursor”), which we view as a compelling extension of our strategy

to vertically integrate compute infrastructure, models, and applications. Under the compute agreement, we will

provide Cursor with certain GPU cluster compute capacity and collaborate to improve existing models, including

Grok, and potentially to jointly develop AI models and related model-specific deliverables or products. With the

option agreement, we have the right, but not obligation, to acquire Cursor at a predetermined price or pay a fee. We

13

consider software development as a strategically important use case for AI given its combination of high-quality

structured data, rapid feedback cycles and frequent, mission-critical usage. AI-assisted coding workflows generate

context-rich, verifiable data that can enhance model training and performance, while also driving sustained inference

demand. The depth of Cursor’s integration with a high-frequency coding workflow generates valuable developer

interaction data, including coding generation prompts, iteration cycles, and software architecture decisions. We

expect that access to this data will enhance our model training and inference, including with respect to Grok.

Meanwhile, by providing access to our large-scale compute infrastructure, we believe we can help Cursor deliver

faster and higher quality user experiences. The collaboration with Cursor may also accelerate our AI strategy by

integrating our AI models more directly into developer workflows and expanding the distribution of our AI

capabilities through high-engagement software interfaces.

The consideration for the acquisition of Cursor, if any, after the closing of this offering would consist of shares of

our Class A common stock based on an implied equity value of Cursor of $60.0 billion, and the price of our Class A

common stock that equals the volume-weighted average closing price thereof over the seven consecutive trading

days immediately preceding the closing of the acquisition. If either (i) we decide to terminate the option agreement

or (ii) Cursor is eligible to and decides to terminate due to our material breach of the option agreement (subject to

notice and cure provisions), Cursor is entitled to a $1.5 billion termination fee under the option agreement and an

$8.5 billion deferred services fee under the compute agreement. These fees are payable in cash (or Class A common

stock, if this offering has not been consummated at the time the fees become payable). For more information about

our arrangement with Cursor, including our option to acquire the company, please refer to “Business—Collaboration

with Cursor” included elsewhere in this prospectus.

Compute Services Agreements with Third Parties

We believe our compute infrastructure and related strategy provides us with substantial flexibility in how we

allocate and monetize capacity. We have the ability to use compute resources to support our proprietary AI

applications (such as Grok 5, which is currently being trained at COLOSSUS II), while also providing access to

select compute capacity to third-party customers. For example, in May 2026, we entered into Cloud Services

Agreements with Anthropic PBC (“Anthropic”), an AI research and development public benefit corporation, with

respect to access to compute capacity across COLOSSUS and COLOSSUS II. Pursuant to these agreements, the

customer has agreed to pay us $1.25 billion per month through May 2029, with capacity ramping in May and June

2026 at a reduced fee. The agreements may be terminated by either party upon 90 days’ notice. The customer will

retain ownership and intellectual property rights in its content, AI models, and related data. This structure allows us

to monetize unused compute capacity in our infrastructure, while still permitting reallocation of the capacity for our

own internal initiatives if needed in the future. We have sufficient capacity to provide compute for our own AI

models, including support of our training and inference demands, and to satisfy the obligations under these

agreements. We expect to enter into additional similar services contracts. We believe this opportunity highlights the

increasing importance of large-scale, frontier-level AI infrastructure and positions us as a differentiated provider of

high-performance compute capacity to both internal and third-party AI workloads. We believe our dual monetization

strategy provides multiple pathways to generate returns on invested capital.

Founder, Chief Executive Officer, Chief Technical Officer and Chairman of Our Board

Mr. Musk is our founder, Chief Executive Officer, Chief Technical Officer and the Chairman of our board.

Assuming a size as set forth on the cover page of this prospectus and an initial public offering price of $               

per share (the midpoint of the estimated price range set forth on the cover page of this prospectus), Mr. Musk will

hold approximately           % of the voting power of our common stock (or               % if the underwriters exercise

their option to purchase additional shares of Class A common stock in full) immediately after this offering through

his ownership of                    shares of our Class A common stock and                    shares of our Class B common

stock, which comprises approximately           % of our Class B common stock. Under our charter, the holders of our

Class B common stock will have the right to elect a majority of our board (such directors, the “Class B Directors”),

for so long as any shares of Class B common stock remain outstanding. As the holder of a majority of our shares of

Class B common stock, Mr. Musk will be able to elect, remove or fill any vacancy among the Class B Directors. In

addition, for so long as he beneficially owns more than 50% of the voting power of our common stock, Mr. Musk

will control the voting power over the selection of our board. As a result, Mr. Musk will have the power to control

14

the outcome of matters requiring shareholder approval, including election of all our directors, and to control our

business and affairs.

Our Controlled Company Status

We will be a controlled company as of the completion of this offering under Nasdaq and Nasdaq Texas listing rules.

A controlled company is not required to have a majority of its board composed of independent directors or to

establish independent compensation and nominating committees. As a controlled company, we will remain subject

to rules that require us to have an audit committee composed entirely of independent directors.

Corporate Information

We were founded and incorporated as Space Exploration Technologies Corp., a Delaware corporation, on March 14,

2002 and reincorporated as a Texas corporation on February 14, 2024. Our principal executive offices are located at

1 Rocket Road, Starbase, Texas 78521. Our website address is www.spacex.com. Information contained on our

website or linked therein or otherwise connected thereto does not constitute part of nor is it incorporated by

reference into this prospectus or the registration statement of which this prospectus forms a part.

15

Summary of Risk Factors

An investment in our Class A common stock involves risks and uncertainties. The following is a summary of the

principal factors that make an investment in our Class A common stock speculative or risky, all of which are more

fully described below in the section titled “Risk Factors.” This summary should be read in conjunction with the

“Risk Factors” section and should not be relied upon as an exhaustive summary.

•Any failure or delay in the development of Starship at scale or in achieving the required launch cadence,

reusability and capabilities thereafter would delay or limit our ability to execute our growth strategy, including

the deployment of next-generation satellites, global satellite-to-mobile connectivity, and orbital AI compute,

which could materially adversely affect our business, financial condition, results of operations, and future

prospects.

•Any delays or difficulties in obtaining, maintaining or renewing required regulatory approvals and licenses

required for our space-related activities, including the U.S. Federal Aviation Administration (“FAA”) launch

and reentry licenses, would materially delay or disrupt our operations, harm our business, or limit our ability to

execute our business strategy.

•Any delays or difficulties in obtaining, maintaining or renewing required communications licenses and

spectrum authorizations for our satellite connectivity services, including international and FCC satellite

spectrum licenses, could materially delay or disrupt our operations, harm our business, or limit our ability to

execute our business strategy.

•Our AI products and X platform are subject to complex and evolving U.S. and foreign laws and regulations that

are subject to change and uncertain interpretation, and we could be required to make changes to our products

and business practices, and be exposed to monetary penalties, increased cost of operations, declines in user

growth or engagement, or loss of customers, or other harm to our AI products and X platform.

•Our business strategy depends on successfully designing, developing, and deploying our products and services,

as well as related platforms, infrastructure, and other strategic initiatives, at an unprecedented scale, which

presents significant execution, cost, and timing risks.

•We have experienced, and will likely continue to experience, launch delays and failures that could have a

material adverse effect on our business, financial condition, results of operations, and future prospects.

•Our satellites, launch vehicles, and other space-related technologies operate, and in the case of orbital AI

compute, will operate, in the harsh and unpredictable environment of space, exposing them to a wide and

unique range of space-related risks that could cause them to malfunction or fail, and any such malfunction or

failure could adversely affect our business, financial condition, results of operations, and future prospects.

•The continued proliferation of satellite constellations in Low-Earth Orbit, as well as the risk of collisions with

space debris or other spacecraft, could limit or impair our launch flexibility and satellite deployment, which

could adversely affect our business, financial condition, results of operations, and future prospects.

•Interruptions in the operation of critical satellite network, ground station, launch, manufacturing, or spacecraft

or data center infrastructure could result in significant downtime, operational delays or loss of service, each of

which could have a material adverse effect on our business, financial condition, results of operations, and future

prospects.

•Manufacturing, testing and launching rockets, satellites, and spacecraft, including our efforts to reuse rockets

and spacecraft, involve inherent risks that could result in human injury or death, property damage and

environmental damage or other adverse environmental impacts due to accidents or equipment failures. Any such

events could result in substantial losses, including reputational harm and legal liability, which could have a

material adverse effect on our business.

•Although we are focused on the vertical integration of our businesses, we depend on third parties to

manufacture and supply certain key components necessary for the provision of our launch, connectivity, and AI

16

services, and any supply shortages or disruptions or failures in their performance could have a material adverse

effect on our business, financial condition, results of operations, and future prospects.

•Our ability to scale our AI products relies on our terrestrial and orbital AI compute infrastructure, which

depends on the availability of power, AI processors, and other critical components, telecommunications

services, and any shortages or disruptions thereof would materially adversely affect our business, financial

condition, results of operations, and future prospects.

•We face intense competition in the markets in which we operate, and while we have historically outperformed

certain competitors in our Space and Connectivity segments, we may not continue to do so, which could

adversely affect our business, financial condition, results of operations, and future prospects.

•The Company’s AI segment is recently formed, still being integrated, operates in a rapidly evolving industry

and is subject to integration, execution, competitive and operational risks.

•Adverse global macroeconomic and geopolitical conditions may negatively affect our business, financial

condition, results of operations and future prospects.

•We depend on our ability to recruit and retain employees who have advanced engineering and technical skills,

and intense competition for such employees may increase costs and affect our ability to meet development and

production timelines.

•Any significant disruption in, or unauthorized access to, our computer and data systems or those of third parties

that we utilize in our operations could result in a loss or degradation of service, loss of trust in us and harm to

our business.

•The development and maintenance of the technologies and infrastructure necessary to support our current and

future operations will require significant capital expenditures, and if we are unable to generate sufficient cash

flow from operations or obtain additional financing on acceptable terms, our business, financial condition,

results of operations, and future prospects could be materially and adversely affected.

•Our substantial level of indebtedness could materially adversely affect our financial condition.

•Our future revenue and operating results depend upon our ability to develop new technologies and respond to

changes in customer demands and industry standards in highly competitive markets, and if we are unable to do

so, our business, financial condition, results of operations, and future prospects may be materially and adversely

affected.

•The estimates of future market opportunity and forecasts of market growth, and our ability to capture such

markets, included in this prospectus may prove to be inaccurate.

•Many of our initiatives, including those to develop orbital AI compute at scale, manufacture AI chips at scale,

establish a lunar economy, develop human augmentation systems, and transport humans and cargo to the Moon

and Mars, involve significant technical complexity, unproven technologies, or technologies that do not exist or

may require significant advancement, and such initiatives may not achieve commercial viability.

•The global nature of our business poses risks with respect to unstable, malicious or arbitrary legal regimes and

authorities.

•Our bylaws place restrictions on the forum, venue and procedures for legal actions or proceedings initiated by

our shareholders, including certain requirements for mandatory arbitration. These provisions could limit our

shareholders’ ability to pursue certain claims and/or increase the cost of doing so and could also affect the

procedures, rights, and remedies available to our shareholders in such legal actions or proceedings.

•Upon completion of this offering, Mr. Musk will serve as our Chief Executive Officer, Chief Technical Officer,

and Chairman of our board and control the election of our directors, and our dual class structure concentrates

voting control with Mr. Musk and other holders of our Class B common stock. This will limit or preclude your

ability to influence corporate matters and the election of our directors.

17

The Offering

Issuer......................................................................	Space Exploration Technologies Corp.
Class A common stock offered by us.....................	shares (or                shares if the underwriters exercise their option to purchase additional shares of Class A common stock in full).
Class A common stock outstanding immediately after this offering................................................	shares (or                shares if the underwriters exercise their option to purchase additional shares of Class A common stock in full).
Class B common stock outstanding immediately after this offering................................................	shares.
Voting power of Class A common stock after giving effect to this offering...............................	% (or                % if the underwriters exercise their option to purchase additional shares of Class A common stock in full).
Voting power of Class B common stock after giving effect to this offering...............................	% (or                % if the underwriters exercise their option to purchase additional shares of Class A common stock in full).
Voting rights...........................................................	Each share of Class A common stock will entitle its holder to one vote per share. Each share of Class B common stock will entitle its holder to 10 votes per share. Class A shareholders and Class B shareholders will vote together as a single class on all matters to be voted on by shareholders under our charter, except the holders of our Class B common stock will have the right to elect a majority of our board and have certain other voting rights as a class. Each share of Class B common stock will be convertible at any time at the option of the holder into one share of our Class A common stock. In addition, each share of Class B common stock will convert automatically into one share of Class A common stock upon a Transfer (as defined in the charter) of that share of Class B common stock, whether or not for value, except for Permitted Transfers (as defined in the charter). Please refer to “Description of Capital Stock.”
Use of proceeds......................................................	We expect to receive approximately $                of net proceeds from this offering (or $                if the underwriters exercise their option to purchase additional shares of Class A common stock in full), based upon the assumed initial public offering price of $                per share (which is the midpoint of the price range set forth on the cover page of this prospectus), after deducting underwriting discounts and commissions and estimated offering expenses payable by us. Please refer to “Underwriting.” We intend to use the net proceeds from this offering to fund our growth strategy, including the expansion of our AI compute infrastructure, enhancements to our launch infrastructure and launch vehicles, increases in the scale and capacity of our satellite constellations, and any remaining amounts for general corporate purposes. Please refer to “Use of Proceeds” for a more complete description of the intended use of proceeds from this offering.

18

Dividend policy......................................................	We do not anticipate declaring or paying any cash dividends to holders of our common stock in the foreseeable future. We currently intend to retain future earnings, if any, to finance the growth of our business. Our future dividend policy is within the discretion of our board and will depend upon then-existing conditions, including our results of operations, financial condition, capital requirements, investment opportunities, statutory restrictions on our ability to pay dividends, restrictions in our existing and any future debt agreements and other factors our board may deem relevant. Covenants under our Credit Agreements also restrict our ability to pay dividends, and we may enter into credit agreements or other borrowing arrangements in the future that restrict our ability to declare or pay cash dividends or make distributions in the future.
Directed share program..........................................	At our request, the underwriters have reserved              percent of the shares of Class A common stock to be issued by the Company and offered by this prospectus for sale, at the initial public offering price, to employees of the Company and certain other designated individuals. If purchased by these persons, these shares of Class A common stock will not be subject to a lock-up restriction. The number of shares of Class A common stock available for sale to the general public will be reduced to the extent these individuals purchase such reserved shares of Class A common stock. Any reserved shares of Class A common stock that are not so purchased will be offered by the underwriters to the general public on the same basis as the other shares of Class A common stock offered by this prospectus.
Controlled company...............................................	Upon completion of this offering, Mr. Musk will beneficially own a majority of the voting power of our common stock and the Class B common stock, which elects a majority of the board. As a result, we expect to be a “controlled company” within the meaning of the Nasdaq and Nasdaq Texas corporate governance standards, and intend to rely on exemptions from certain of the corporate governance listing requirements. Please refer to “Management—Controlled Company Exemption” and “Certain Relationships and Related Person Transactions.”
Risk factors.............................................................	You should carefully read and consider the information set forth in the section titled “Risk Factors” beginning on page 26, together with all of the other information set forth in this prospectus, before deciding whether to invest in our Class A common stock.
Listing and trading symbol.....................................	We have applied to list our Class A common stock on Nasdaq and Nasdaq Texas under the symbol “SPCX.”

The number of shares of our Class A and Class B common stock that will be outstanding after this offering is based

on           shares of Class A common stock and                   shares of Class B common stock outstanding as of March

31, 2026, after giving effect to (i) the sale of                    shares of Class A common stock in this offering, (ii) the

Class C Reclassification (as defined below), and (iii) the Preferred Conversion (as defined below).

Unless otherwise noted, common stock outstanding after the offering and other information based thereon in this

prospectus does not reflect any of the following:

•                shares of Class A common stock issuable upon exercise of the underwriters’ option to purchase

additional shares from us;

•                shares of Class A common stock issuable upon the exercise of outstanding stock options granted

under the Equity Plans (as defined below) that were outstanding as of March 31, 2026 with a weighted-average

exercise price of $                per share;

19

•                shares of Class A common stock issuable upon the exercise of outstanding stock options granted

under the Equity Plans (as defined below) granted after March 31, 2026 with a weighted-average exercise price

of $                per share;

•                shares of Class A common stock issuable upon the vesting and settlement of restricted stock units that

were outstanding as of March 31, 2026 under the Equity Plans (none of which will vest in connection with this

offering);

•                shares of Class A common stock issuable upon the vesting and settlement of restricted stock units

granted after March 31, 2026 under the Equity Plans (none of which will vest in connection with this offering);

•                shares of Class A common stock reserved for issuance under our Amended and Restated 2024 Equity

Incentive Plan (the “A&R 2024 Plan”), excluding shares subject to outstanding awards thereunder as described

above, which we plan to adopt in connection with this offering;

•                shares of Class A common stock reserved for issuance under our Amended and Restated 2017 Equity

Stock Purchase Plan (the “A&R 2017 ESPP”), which we plan to adopt in connection with this offering; and

•                shares of Class A common stock reserved for future issuance upon the conversion of           shares of

Class B common stock on a one-for-one basis.

The term “Equity Plans” refers to our 2015 Plan, our A&R 2017 ESPP and our A&R 2024 Plan as well as (i) xAI’s

2023 Equity Incentive Plan, 2023 Incentive Plan and 2025 Equity Incentive Plan, each of which we assumed in the

xAI Merger and (ii) the 2017 Stock Plan, as amended, of Swarm Technologies, Inc. (“Swarm”), which we assumed

in our acquisition of Swarm in 2021.

The information in this prospectus also does not reflect:

•the payment of                      shares of Class A common stock and cash consideration which would occur upon

closing of our agreement with EchoStar Corporation (“EchoStar”) to purchase certain AWS-3, AWS-4, and H-

Block spectrum licenses pursuant to the License Purchase Agreement, dated as of September 7, 2025 (as

amended and restated on November 5, 2025), by and among SpaceX, Spectrum Business Trust 2025-1 and

EchoStar (the “Spectrum Transaction”), which transaction was approved by the FCC on May 12, 2026 and is

subject to other closing conditions prior to completion; and

•the issuance of shares of our Class A common stock if in the future our board determines to exercise our option

to acquire Cursor as such option is described under “Business—Collaboration with Cursor,” which, as an

example, assuming the volume-weighted average closing price of our common stock over the seven consecutive

trading days immediately preceding the closing of such acquisition were equal to the initial public offering price

of $per share (which is the midpoint of the price range set forth on the cover of this prospectus),

would equal approximatelyshares. The actual number of shares that may be issued will be

determined based on a future trading price and is subject to customary adjustments for reclassifications,

recapitalization, stock splits or any other similar event affecting the outstanding capital stock of Cursor or the

Company.

Unless otherwise indicated, all information contained in this prospectus assumes or gives effect to:

•the 2026 Stock Split;

•prior to the completion of this offering, pursuant to the terms of our certificate of formation in effect as a private

company prior to this offering, the reclassification of all of the outstanding shares of our Class C common stock

into an aggregate of                      shares of Class A common stock (the “Class C Reclassification”) and the

conversion of the outstanding shares of all our preferred stock into an aggregate of                     shares of our

Class A common stock and                    shares of our Class B common stock (the “Preferred Conversion”);

•the effectiveness of our charter and bylaws, which will become effective upon the completion of this offering;

20

•an initial public offering price of $                per share of Class A common stock (the midpoint of the price range

set forth on the cover of this prospectus);

•that the underwriters do not exercise their option to purchase additional shares of Class A common stock from

us; and

•no purchase of shares of Class A common stock in this offering by our directors, officers or existing

shareholders.

21

Summary Historical Consolidated Financial and Operating Data

The following table sets forth the summary historical consolidated financial and operating data for the periods and as

of the dates presented. The summary historical consolidated financial data as of March 31, 2026 and for the three

months ended March 31, 2026 and 2025 (except for pro forma basic and diluted net loss per share of common stock

attributable to common shareholders and weighted average shares used in computing pro forma basic and diluted net

loss per share of common stock attributable to common shareholders) has been derived from our unaudited

consolidated financial statements included elsewhere in this prospectus. The summary historical consolidated

financial data as of December 31, 2025 and 2024 and for the years ended December 31, 2025, 2024, and 2023

(except for pro forma basic and diluted net loss per share of common stock attributable to common shareholders and

weighted average shares used in computing pro forma basic and diluted net loss per share of common stock

attributable to common shareholders) has been derived from our audited consolidated financial statements included

elsewhere in this prospectus.  The summary historical consolidated financial and operating data presented below is

not indicative of the results to be expected for any future period, and the results for any interim period are not

necessarily indicative of the results to be expected for the full fiscal period.

The summary historical consolidated financial and operating data of SpaceX has been prepared to reflect the

retrospective combination of the companiesfor all periods presented to include the historical results of xAI, which

was acquired by SpaceX, effective February 2, 2026, and X Holdings, which was acquired by xAI, effective

March 28, 2025, because these transactions were between entities under common control.

The following information should be read together with “Management’s Discussion and Analysis of Financial

Condition and Results of Operations” and our consolidated financial statements and related notes thereto included

elsewhere in this prospectus. The summary historical consolidated financial data included in this section is not

intended to replace the consolidated financial statements and is qualified in its entirety by our consolidated financial

statements and related notes included elsewhere in this prospectus.

Statements of Operations Data:

	Three Months Ended March 31,		Year Ended December 31,
	2026	2025	2025	2024	2023
(in millions, except per share data)	(unaudited)
Revenue .............................................	$4,694	$4,067	$18,674	$14,015	$10,387
Total costs and expenses...........	6,637	4,040	21,263	13,549	13,892
Income (loss) from operations ...........	(1,943)	27	(2,589)	466	(3,505)
Net income (loss) ..............................	$(4,276)	$(528)	$(4,937)	$791	$(4,628)
Net income (loss) per share of common stock attributable to common shareholders (1)
Basic..............................................	$(1.27)	$(0.18)	$(1.69)	$0.01	$(1.68)
Diluted...........................................	$(1.27)	$(0.18)	$(1.69)	$0.00	$(1.68)
Weighted average shares used in computing net income (loss) per share of common stock (1)
Basic..............................................	3,884	2,875	2,926	2,848	2,759
Diluted...........................................	3,884	2,875	2,926	9,956	2,759

__________________

(1)Please refer to Note 14, Earnings per Share to our audited consolidated financial statements appearing elsewhere in this prospectus for an

explanation of our calculation of basic and diluted net income (loss) per share of common stock attributable to common shareholders.

22

The following table sets forth the computation of unaudited pro forma basic and diluted net loss per share of

common stock attributable to common shareholders for the period presented:

(in millions, except per share data)	Three Months Ended March 31, 2026	Year Ended December 31, 2025
Numerator:
Net loss attributable to common shareholders, basic and diluted.........................	$(4,947)	$(4,937)
Pro forma adjustment to reverse the deemed dividend on SpaceX Redeemable Convertible Preferred Stock, basic and diluted ................................................	565	—
Pro Forma net loss attributable to common shareholders, basic and diluted........	$(4,382)	$(4,937)
Denominator:
Weighted average shares used in computing net loss per share of common stock, basic and diluted.....................................................................................	3,884	2,926
Pro forma adjustment to reflect the Preferred Conversion as if the conversion occurred on January 1, 2025, basic and diluted................................................	6,723	6,723
Weighted average shares used in computing pro forma net loss per share of common stock, basic and diluted......................................................................	10,607	9,649
Pro forma net loss per share of common stock attributable to common shareholders, basic and diluted (2)..........................................................................	$(0.41)	$(0.51)

__________________

(2)Pro forma basic and diluted net loss per share of common stock attributable to common shareholders and weighted-average number of

shares used in the computation of the per share amount gives effect to (i) the Preferred Conversion as if such conversion had occurred as of

January 1, 2025, (ii) the Class C Reclassification as if such reclassification had occurred as of January 1, 2025, and (iii) the effectiveness of

our charter, which will become effective upon the completion of this offering.

Statement of Cash Flows Data:

	Three Months Ended March 31,		Year Ended December 31,
	2026	2025	2025	2024	2023
(in millions)	(unaudited)
Net cash provided by operating activities..........................................	$1,047	$727	$6,785	$5,776	$4,520
Net cash used in investing activities..	$(16,724)	$(4,170)	$(19,575)	$(10,796)	$(4,867)
Net cash provided by financing activities..........................................	$7,125	$354	$26,350	$11,830	$422

Capital Expenditures:

The following table presents our capital expenditures by segment:

	Three Months Ended March 31,		Year Ended December 31,
	2026	2025	2025	2024	2023
(in millions)	(unaudited)
Space..................................................	$1,052	$759	$3,832	$2,032	$1,497
Connectivity.......................................	1,332	814	4,178	3,498	2,455
AI........................................................	7,723	2,567	12,727	5,633	463
Total Capital Expenditures.................	$10,107	$4,140	$20,737	$11,163	$4,415

23

Balance Sheet Data:

	March 31,	December 31,
	2026	2025	2024
(in millions)	(unaudited)
Cash and cash equivalents..............................................................	$15,852	$24,747	$11,385
Total current assets.........................................................................	29,732	30,952	16,108
Property, plant, and equipment, net................................................	53,879	42,602	21,147
Total assets ....................................................................................	102,094	92,079	57,062
Debt and finance leases, current ....................................................	1,538	928	372
Total current liabilities...................................................................	24,436	21,400	11,791
Total liabilities................................................................................	60,512	50,754	31,258
Redeemable convertible preferred stock........................................	7,049	38,752	20,941
Total shareholders’ equity .............................................................	34,533	2,573	4,863

Segment Operating and Financial Data (unaudited)

Space:

	Three Months Ended March 31,		Year Ended December 31,
	2026	2025		2025	2024	2023
Mass to Orbit (in metric tons) (1)........	556	450		2,213	1,699	1,210
Launches (number) (1).........................	40	38		170	138	98
Segment income (loss) from operations (in millions)...................	$(662)	$(70)		$(657)	$21	$(1)
Segment Adjusted EBITDA (in millions) (2)......................................	$(351)	$224		$653	$1,154	$997

Connectivity:

	Three Months Ended March 31,		Year Ended December 31,
	2026	2025		2025	2024	2023
Starlink Subscribers (in millions) (1)...	10.3	5.0		8.9	4.4	2.3
Starlink ARPU (dollars per month) (1)	$66	$86		$81	$91	$99
Segment income from operations (in millions)..........................................	$1,188	$1,033		$4,423	$2,006	$469
Segment Adjusted EBITDA (in millions) (2)......................................	$2,087	$1,618		$7,168	$3,849	$1,602

AI:

	Three Months Ended March 31,		Year Ended December 31,
	2026	2025		2025	2024	2023
Nameplate compute draw (in gigawatts) (1)....................................	1	0.3		0.8	0.3	0
Segment loss from operations (in millions)..........................................	$(2,469)	$(936)		$(6,355)	$(1,561)	$(3,973)
Segment Adjusted EBITDA (in millions) (2)......................................	$(609)	$(112)		$(1,237)	$347	$1,222

______________

(1)Please refer to the section titled “Management’s Discussion and Analysis of Financial Condition and Results of Operation—Key Business

Metrics” for additional information on our key business metrics.

(2)Segment Adjusted EBITDA is a non-GAAP measure. Please refer to the section titled “Management’s Discussion and Analysis of Financial

Condition and Results of Operation—Non-GAAP Financial Measures” for additional information on our non-GAAP financial measures,

including reconciliations of Segment Adjusted EBITDA to segment income (loss) from operations, the most directly comparable GAAP

measure.

26

RISK FACTORS

Investing in our Class A common stock involves a high degree of risk. You should carefully consider the risks and

uncertainties described below, together with all of the other information contained in this prospectus, including our

consolidated financial statements and the related notes thereto, before making a decision to invest in our Class A

common stock. The disclosures in this section reflect our beliefs and opinions as to factors that could materially and

adversely affect us in the future. We may not be able to accurately predict, control, or mitigate these risks.

References to past events are provided by way of example only and are not intended to be a complete listing or a

representation as to whether or not such factors have or have not occurred in the past or their likelihood of

occurring in the future. Additional risks and uncertainties that we are unaware of, or that we currently believe are

not material, may also become important factors that adversely affect us. Many of the risks and uncertainties that

could materially adversely affect us or our prospects are beyond our control or relate to portions of our business

strategy that have a lengthy time horizon or involve unprecedented ventures. This can make assessment of certain

risks more difficult and you should factor these uncertainties into your assessment of an investment in our Class A

common stock. If any of the following risks and uncertainties occur, the price of our Class A common stock could

decline, and you could lose part or all of your investment.

Risks Related to Our Business

Any failure or delay in the development of Starship at scale or in achieving the required launch cadence,

reusability and capabilities thereafter would delay or limit our ability to execute our growth strategy, including

the deployment of next-generation satellites, global satellite-to-mobile connectivity, and orbital AI compute,

which could materially adversely affect our business, financial condition, results of operations, and future

prospects.

If we are unable to successfully complete the development, testing, and deployment of Starship at scale in

accordance with our anticipated schedule, or at all, or if we are unable to achieve sufficient launch cadence,

reusability, and capability, our ability to execute our growth strategy (such as the deployment of our next-generation

V3 satellites, V2 satellite-to-mobile connectivity, and providing orbital AI compute infrastructure) would be

materially and adversely affected. The commercial deployment of Starship, particularly at scale, is subject to

substantial risks and uncertainties inherent in the development of new and complex technologies and systems.

Delays or challenges in the Starship program have in the past occurred, and may occur in the future due to a variety

of factors, including unforeseen technical challenges, supply chain disruptions, manufacturing difficulties, delays in

the development, construction or commissioning of launch and fueling infrastructure (such as launch pads, air

separation units and other propellant production systems), unavailability of such launch and fueling infrastructure

(including launch pads) in sufficient number and in operable condition (including as a result of mishaps), loss or

damage to the vehicle or other components, regulatory hurdles, or the need for additional design modifications. If we

are required to undertake unanticipated redesigns, conduct additional testing, replace lost vehicles or components, or

address operational setbacks, we may experience delays and incur significant additional costs, or be forced to

reallocate critical resources from other projects. If our launch pads are not available for an extended period of time

for any reason, we may not be able to achieve our development, testing and deployment goals. Such delays could

have cascading effects on our ability to achieve the scale we need to timely achieve future objectives. In addition, a

critical part of our growth strategy involves increasing our launch cadence, reusability and capability, including

increasing our payload per launch. This will require, among other things, the successful development and operation

of reusable launch vehicles, substantially increased access to raw materials and components like steel, fuel and

propellant, the construction of additional facilities and securing of additional launch sites or rights to additional

launches from existing sites, and navigating complex and evolving regulatory requirements and environmental and

technological issues as we seek to increase our launch cadence. Our rocket programs have historically required

substantial time and resources to reach the cadence and cost thresholds necessary for commercial viability, and the

development of Starship may face similar or greater challenges. Any significant delay in achieving key development

milestones, obtaining the necessary regulatory approvals or increasing and maintaining our launch cadence,

reusability, and capability would impede the expansion of our service offerings, defer anticipated revenue streams,

and negatively impact our growth trajectory and competitive positioning in rapidly evolving markets.

27

Our ability to execute our growth strategy is highly dependent on Starship. If we are unable to achieve the

commercial development, anticipated performance, launch cadence, or cost efficiencies associated with Starship

within expected timeframes, our ability to deploy next-generation V3 satellites, V2 Mobile satellites, and orbital AI

compute infrastructure at scale, reduce capital and operating costs (including cost per token), realize projected

revenue growth, and retain existing customers from these initiatives could be materially and adversely affected. This

includes our expectations with respect to completion of flight testing of Starship and commencement of payload

delivery to orbit. Our current operational rockets, including Falcon 9 and Falcon Heavy, are not capable of

deploying V3 satellites and V2 Mobile satellites.

In addition, our ability to pursue new initiatives and capture emerging business opportunities—particularly those

requiring high launch cadence, large payload capacity, or advanced in-space capabilities, such as lunar operations

and interplanetary missions—depends on the timely and successful deployment of Starship and achieving our

targeted launch cadence. Achieving our targeted launch cadence will require significant progress on several key

milestones and the continued investment of significant capital resources. These include: securing additional land and

developing high-rate launch sites and supporting infrastructure across multiple locations; scaling production of

Starship vehicles and Raptor engines; constructing propellant production facilities, including air separation units and

methane liquefaction plants co-located with launch sites; securing sufficient power supply; and obtaining the

necessary regulatory approvals, particularly from the FAA, to support a high launch cadence while addressing public

safety and environmental considerations. We face a number of material challenges and uncertainties in achieving

these milestones, such as achieving reliable high-cadence return-to-launch-site operations for the full vehicle stack,

developing durable reusable heat shields capable of withstanding repeated high-velocity reentries, ensuring rapid

refurbishment and high-rate reusability of engines and other vehicle components, managing public and regulatory

tolerance for anomalies during the transition to frequent operational flights, securing sufficient power for both

manufacturing and launch operations, and obtaining timely regulatory approvals from the FAA and other agencies.

Orbital refueling involves technical complexities associated with cryogenic propellant transfer in microgravity,

propellant settling, and boil-off management and is required for lunar and interplanetary objectives.

If Starship does not achieve full reusability or rapid turnaround, we may experience higher per-launch costs, slower

deployment timelines for our large-scale constellations (including our orbital AI compute program), delayed revenue

growth, and increased overall capital requirements, and our brand and reputation may suffer. AI compute satellites at

scale need full Starship reusability to be economically compelling. Without full reusability and rapid turnaround,

Starship would still be capable of enabling progress on our next-generation Starlink, direct-to-cell, initial lunar

objectives, and early AI compute satellite deployments, but such progress would be at a slower pace and higher cost.

Any inability to deliver Starship to market as planned could constrain our participation in new or expanding

addressable markets, limit our competitive differentiation, and hinder our efforts to attract and retain customers.

There can be no assurance that we will be able to achieve our objectives with respect to Starship within the expected

timeframes, if at all, or that delays or setbacks will not materially impact our strategic plans.

Any delays or difficulties in obtaining, maintaining or renewing required regulatory approvals and licenses

required for our space-related activities, including FAA launch and reentry licenses, would materially delay or

disrupt our operations, harm our business, or limit our ability to execute our business strategy.

Our launch services are subject to extensive regulation in the United States and internationally. We must secure and

maintain numerous governmental approvals to launch our rockets and conduct related launch and reentry activities.

Any failure or significant delay in obtaining required licenses and permits or failure to maintain them could disrupt

our operations, constrain our growth, and adversely affect our ability to serve our customers. Our plans to deploy

large-scale orbital infrastructure, including orbital AI compute systems, will require the operation of very large

satellite constellations, potentially numbering up to one million satellites. These plans will depend on obtaining a

wide range of domestic and international approvals, including spectrum authorizations, orbital debris mitigation

approvals, and coordination and authorization requirements relating to space situational awareness and international

regulatory regimes, and there can be no assurance that such approvals will be obtained on acceptable timelines,

terms, or at all.

We depend on timely approvals from the FAA to conduct our launch operations. If we do not receive FAA launch

licenses or related approvals on the schedules we anticipate or if we are subject to regulatory delays, we could be

28

forced to delay or cancel planned launches, which could cause missed customer commitments, increased costs, and

underutilization of our launch resources. Obtaining a launch license involves rigorous safety and environmental

reviews, and unforeseen issues in meeting these requirements or additional conditions imposed during the review

process could also impact our launch timelines. For example, current FAA regulations do not permit return-to-

launch-site reentries for Starship, requiring us to obtain a waiver from the FAA, which is not guaranteed and could

delay or restrict such operations. Following an anomaly, mishap, or failure, the FAA or other authorities may require

investigations, impose corrective actions, or restrict or delay our ability to conduct launch operations. We have in the

past been, and may in the future become, subject to such actions, impacting our ability to increase launch cadence.

The regulatory framework governing commercial launches may also evolve over time. The FAA or other authorities

could introduce new or more stringent requirements for launch licensing – for instance, heightened safety standards,

environmental mitigation measures, or other operational restrictions – that could require us to invest in new

technologies, adjust our procedures, or otherwise incur additional compliance burdens. Moreover, as the frequency

of our launches and industry activity overall continues to grow, the FAA’s resources may become strained, which

could lead to longer application processing times and other difficulties obtaining FAA licenses. Any significant

delay in receiving required FAA licenses, the imposition of onerous new licensing conditions, or failure to obtain an

approval for a key launch, could materially adversely affect our business, financial condition, results of operations,

and future prospects.

Any delays or difficulties in obtaining, maintaining or renewing required communications licenses and spectrum

authorizations for our satellite connectivity services, including international and FCC satellite spectrum licenses,

could materially delay or disrupt our operations, harm our business, or limit our ability to execute our business

strategy.

Our satellite connectivity services are subject to extensive regulation in the United States and internationally.

Obtaining and maintaining communications licenses and approvals from U.S. and foreign regulatory authorities is

critical to our connectivity services. Our satellite connectivity, including our global satellite-to-mobile connectivity

services under Starlink Mobile, depend on access to radio frequency spectrum and authorizations from the FCC in

the United States and telecommunications regulators in other countries. Without these licenses and approvals, we

generally cannot offer connectivity services in a given market. Acquiring the necessary authorizations can be a

complex and time-consuming process, often involving technical coordination, public-interest or national security

reviews, and cross-border considerations, including in certain jurisdictions where regulatory processes may be

influenced by protectionist policies or preferences. Spectrum access itself is limited and highly regulated. In

September 2025, we announced a definitive agreement with EchoStar to purchase its AWS-4 and H-block spectrum

licenses. The Spectrum Transaction was approved by the FCC on May 12, 2026 and is subject to other closing

conditions prior to completion. We expect the Spectrum Transaction to close in November 2027. There can be no

assurance that these conditions will be satisfied or waived in a timely manner, or at all. Even if the transaction is

completed, there can be no assurance that our purchase of licenses from EchoStar will be sufficient to meet our

growing need for spectrum licenses and we may be unable to find other parties to provide us with additional

spectrum licenses on terms acceptable to us, or at all. We may in the future pursue additional acquisitions, leases, or

other arrangements relating to spectrum rights in order to support the expansion of our connectivity services, and

there can be no assurance that we will be able to enter into or complete any such transactions or arrangements on

acceptable terms, or at all. Any such future transactions or arrangements could require significant capital

commitments, ongoing payment obligations, and regulatory approvals. In addition, we must secure the global right

to use the spectrum acquired from EchoStar from a number of international telecommunications regulators in order

to make our V2 satellite-to-mobile services usable worldwide, and there can be no assurance that such authorizations

will be granted on acceptable terms, or at all. Moreover, our rights to use certain frequencies are coordinated through

the International Telecommunication Union (“ITU”) and are subject to international agreements to prevent harmful

interference. We must comply with ITU rules and coordination procedures, and changes in international spectrum

allocations or adverse decisions in global regulatory forums could also reduce the frequencies available to us or

attach conditions that degrade our network’s performance. Additionally, third parties have in the past, and may in

the future, obtain spectrum rights for the purpose of blocking market entry.

Regulatory regimes for communications services vary widely across different countries and are continuously

evolving. Each country may impose its own licensing conditions and operating requirements on satellite internet

29

providers – for example, mandates to partner with a local entity, to host certain infrastructure within its borders, or to

adhere to specific standards relating to data privacy and cybersecurity (including data localization) and, in some

cases, regulators may deny, delay or decline to grant authorization for us to operate or use our spectrum in their

jurisdiction at all. Regimes in certain of our target markets may also favor incumbent or legacy telecommunications

companies, which may impede, delay, or prevent our ability to enter such markets. Compliance with the different

requirements of applicable regulatory regimes can be challenging and costly, and any failure to comply with local

laws and regulations could lead to penalties or the loss of our authorization to operate in that region. Furthermore,

communications regulatory authorizations often require periodic renewal and ongoing compliance with conditions

such as deployment milestones, fee payments, and interference mitigation obligations. If we are unable to obtain,

retain, and renew the necessary spectrum rights and service licenses on acceptable terms in each of our target

markets, or if regulatory bodies significantly delay our authorizations or impose burdensome requirements, our

ability to expand and continue our connectivity services would be jeopardized, which would have a material adverse

effect on our business, financial condition, results of operations, and future prospects.

Our AI products and X platform are subject to complex and evolving U.S. and foreign laws and regulations

regarding privacy, cybersecurity, data use, data combination, data protection, content, AI, competition, youth

protection, safety, consumer protection and notification, advertising, e-commerce, sanctions, export controls,and

other matters. Many of these laws and regulations are subject to change and uncertain interpretation, and we

could be required to make changes to our products and business practices, and be exposed to monetary penalties,

increased cost of operations, declines in user growth or engagement, or loss of customers, or other harm to our

AI products and X platform.

Our AI products and X platform are subject to a variety of laws and regulations in the United States and abroad,

including privacy, cybersecurity, data use, data combination, data protection and personal information, the provision

of our services to younger users, biometrics, encryption, rights of publicity and related concepts, content, integrity,

intellectual property, advertising, marketing, distribution, data security, data retention and deletion, data localization

and storage, data disclosure, AI and machine learning, electronic contracts and other communications, competition,

protection of minors, consumer protection, sanctions, export controls, and notification, civil rights, accessibility,

product liability, e-commerce, taxation and online payment services, as well as contractual requirements imposed by

app stores, payment processors, and other partners. The introduction of new products or services, expansion of our

activities in certain jurisdictions, or other actions that we may take may subject us to additional laws, regulations, or

other government scrutiny and, in some cases, such laws, regulations, or government scrutiny may limit or delay our

ability to introduce new products or services or expand our activities in certain jurisdictions. Particularly, our

leadership position in various markets, especially in orbital launch services, could subject us to heightened

regulatory scrutiny under competition laws. In addition, these U.S. and foreign laws and regulations may impose

different obligations from each other. As a result of these laws, regulations, and requirements, we are exposed to the

risk of significant fines and penalties or other adverse consequences, such as changes to our products, services, or

business practices.

Our social media and AI-related activities expose us to a variety of risks related to harmful, misleading or illegal

content, accuracy, misinformation and deepfakes, bias, discrimination, toxicity, sycophancy, AI deception,

consumer protection and notification, products liability, intellectual property infringement or misappropriation,

defamation, data privacy, cybersecurity, and sanctions and export controls. Social media and AI are the subject of

increasing legislative and regulatory activity by various governmental and regulatory agencies in jurisdictions

around the world, which are applying, or are considering applying, platform moderation, intellectual property,

product liability, data privacy, age restrictions, data disclosure, cybersecurity, export controls, consumer protection,

or other existing laws and regulations or new general legal frameworks to AI (such as the EU’s AI Act, California’s

Frontier Artificial Intelligence Act and New York’s Responsible AI Safety and Education Act). In the United States,

an increasing amount of legislative and regulatory activity regarding AI is taking place at the state level. Various

other jurisdictions have enacted or are considering enacting regulations focused on AI. Restrictions under such laws

or regulations, if implemented, could increase the costs and burdens to our AI segment and its customers, delay or

halt deployment of new systems using our AI segment’s products, require us to modify, restrict, or discontinue

certain features (including less constrained modes), and reduce the number of new entrants and customers,

negatively impacting our AI segment’s business and financial results. If we do not adequately address concerns and

30

regulations relating to the responsible use of AI, public confidence in AI could be undermined, adoption of our AI

products and services could slow, and we may suffer reputational or financial harm.

Certain of our AI products, including Grok, offer features or modes designed to generate more candid, direct, or less

reserved or irreverent outputs, such as “Spicy” Imagine Mode and “Unhinged” Voice Mode. These features are

intended to provide users with greater flexibility and control in how they use our tools. Because these modes may be

more irreverent and harsher than our standard offerings, they present heightened risks, including reputational harm,

the generation of potentially explicit content and misinformation or deceptive outputs, potential nonconsensual or

exploitative imagery, intellectual property infringement, or content that could be viewed as exploitative, harmful,

harassing, abusive, or discriminatory. The availability of such features may also increase the risk of regulatory

scrutiny, enforcement actions, litigation, or claims of harm, as well as reputational damage, user or advertiser

backlash, or limitations on our ability to distribute or monetize our products in certain jurisdictions or through

certain partners.

In addition, various regulatory authorities and agencies around the world are actively investigating and making

inquiries relating to social media or the use of AI concerning a variety of matters, including investigations and

inquiries relating to harmful or illegal content, recommendations, advertising, and consumer protection and

notification, which have resulted in, and may in the future result in additional or further investigations and

proceedings being brought against us. Certain features that enable more user-directed or less constrained outputs

may increase the risk of regulatory scrutiny. For example, we are subject to investigations and inquiries from

regulators and law enforcement authorities in the United States and internationally concerning allegations that our

AI products were used to create nonconsensual explicit images or content representing children in sexualized

contexts, and similar matters. We are subject to ongoing litigation, including putative class action lawsuits, relating

to such allegations, and we may be subject to additional litigation in the future concerning these types of allegations.

These regulatory inquiries, including those related to misuse of our AI products, such as Grok, and those related to

the X platform, could expose us to additional investigations, proceedings, and litigation, regulatory sanctions

(including loss of access to certain markets, which has occurred in the past), liability and adverse publicity, any of

which would adversely affect our business.

For example, in February 2026, the Irish Data Protection Commission, our AI segment’s privacy regulator in

Europe, launched a large-scale inquiry to determine whether our AI segment has complied with its obligations under

the European Union’s General Data Protection Regulation (“GDPR”). This inquiry involves the processing of

personal data of European Union data subjects, including children, using generative AI functionality associated with

the Grok model within the X platform. In the United States, the Federal Trade Commission has undertaken an

inquiry into the chatbots of our AI segment and other major technology companies to understand how these

companies have evaluated the safety of their chatbots when acting as companions to children and teens. Regulatory

requirements applicable to online platforms and content moderation, and to AI systems, could require us to

implement costly compliance measures, restrict certain features or jurisdictions, or expose us to significant fines,

liability, penalties, or operational constraints. We are also subject to developer agreements and guidelines imposed

by third-party app stores, such as the Apple App Store and Google Play Store. Failure to comply with these

agreements and guidelines, including those relating to content, could result in the suspension or removal of our

mobile applications from such app stores. Any such suspension or removal could materially limit our ability to

distribute our mobile applications, and adversely affect our business, results of operations, and financial condition.

Authorities around the world have adopted or are considering adopting a number of legislative and regulatory

proposals concerning data protection and privacy. Additionally, the increasing adoption of AI technologies, which

often rely on the collection of large amounts of data and use of such data to train, fine-tune or otherwise develop AI

models, has led data protection authorities around the world to consider and adopt new and evolving interpretations

of data protection laws, imposing specific obligations with respect to the processing of personal data, including

required notices, consents and opt-outs. Adverse legal rulings, legislation or regulations related to such data privacy

matters may result in fines and orders requiring that we change our practices, which could have an adverse effect on

how we provide services, and could harm our business, financial condition, results of operations and future

prospects. These compliance obligations could also cause us to incur substantial costs or harm the quality and

operations of our products and services in ways that harm our business. Further, we are subject to evolving laws and

regulations that dictate whether, how, and under what circumstances we can transfer, receive or otherwise process

31

personal data. The validity of various data transfer mechanisms we currently rely upon remains subject to legal,

regulatory and political developments globally, which may require us to adapt our existing arrangements. Evolving

data protection laws and regulations such as the GDPR and ePrivacy Directive, and regulatory actions affecting our

AI segment may restrict or adversely affect the X platform’s advertising services, Grok’s development and training,

or the ability to offer certain products and services in certain jurisdictions.

We are also subject to tax laws, regulations, and policies of the U.S. federal, state, and local governments and of

comparable taxing authorities in foreign jurisdictions where we conduct business. Changes in tax laws or in their

interpretation or enforcement could result in fluctuations in our effective tax rate, exposure to new or additional tax

liabilities, or adversely affect our after-tax profitability or financial position. These U.S. federal and state, EU, and

other international laws and regulations, which in some cases can be enforced by private parties in addition to

government entities, are constantly evolving and can be subject to significant change. As a result, the application,

interpretation, and enforcement of these laws and regulations are often uncertain, particularly in the new and rapidly

evolving industry in which we operate, and may be interpreted and applied inconsistently from jurisdiction to

jurisdiction and inconsistently with our current policies and practices. For example, regulatory or legislative actions

or litigation concerning the manner in which we display content to our users, moderate content, provide our services

to younger users, or are able to use data in various ways, including for advertising, have in the past and could in the

future adversely affect user growth and engagement, affect the manner in which we provide our services, or

adversely affect our financial results, including by imposing significant fines that increasingly may be calculated

based on global revenue. For example, the UK’s Online Safety Act 2023 and Australia’s Online Safety Amendment

(Social Media Minimum Age) Act 2024 impose risk mitigation and age-related requirements on certain online

platforms. These laws and regulations, as well as any associated claims, inquiries, or investigations or any

government actions, have led to, and may in the future lead to, unfavorable outcomes including increased

compliance costs, changes to our products, loss of revenue, delays or impediments in the development of new

products, negative publicity and reputational harm, increased operating costs, diversion of management time and

attention, and remedies that harm our business, including fines, damages, or orders that we modify or cease existing

business practices. In addition, our AI products and the X platform have historically been, and may continue to be,

subject to claims and investigations relating to misinformation and deepfakes, defamation, intellectual property

infringement or misappropriation, data privacy, cybersecurity, employment matters, advertising practices, and user

harms; defending such matters could be costly and divert management attention.

Our Starlink and other satellite services are subject to complex and evolving U.S. and foreign laws and

regulations, particularly relating to data privacy, cybersecurity, and telecommunications.

Our Starlink and other satellite services are subject to a variety of laws and regulations in the United States and

abroad covering cybersecurity, privacy, data use, data combination, data protection, data security, data retention and

deletion, data localization and storage, and data disclosure to law enforcement agencies. As a satellite internet and

communications provider, we collect and otherwise process various kinds of data in connection with our services,

such as customer personal information, account registration information, device identifiers, network and

connectivity data, and government information. These laws and regulations govern how we handle such information,

and they may, among others, impose requirements relating to cybersecurity and privacy governance, data security

measures, data security breach notification, cross border data transfers, and customer consent obligations.

In particular, the California Consumer Privacy Act (as amended), the GDPR (and its equivalent in the United

Kingdom) and other data privacy laws and regulations impose stringent and burdensome requirements in connection

with the processing of personal information and include significant penalties for non-compliance. Additionally, as a

government contractor, we are also subject to the Department of War’s Cybersecurity Maturity Model Certification

requirements, which requires companies that do business with the Department of War to, depending on the level of

scrutiny required, meet or exceed certain specified cybersecurity standards to be eligible for new contract awards.

Many of these laws and regulations are subject to change and uncertain interpretation, and their application may

vary significantly across jurisdictions. Compliance may require us to modify our policies, procedures, and controls,

and increase our compliance costs and operational complexity. We may post public privacy policies and other

statements regarding our collection, storage, sharing and other processing of personal information, and any actual or

perceived failure to comply with such privacy policies and other statements, as well as the foregoing data privacy

32

and cybersecurity laws and regulations, may subject us to enforcement actions, investigations, litigation, reputational

harm or requirements to modify or cease our business practices.

Our business strategy depends on successfully designing, developing, and deploying our products and services, as

well as related platforms, infrastructure, and other strategic initiatives, at an unprecedented scale, which presents

significant execution, cost, and timing risks.

Our business plan, and ultimately, the achievement of our mission, is predicated on building, commercializing, and

operating products and services, as well as related infrastructure and strategic initiatives at a scale that has not

previously been achieved. This objective requires us to integrate complex technologies, develop new processes and

infrastructure, and coordinate across multiple suppliers, contractors, regulators, and stakeholders. Because we are

attempting to execute at a scale for which there is limited precedent, we face heightened uncertainty with respect to

design, engineering, procurement, construction, commissioning, and operational performance, which is further

heightened by the novel nature of the technologies underlying the products and services we intend to develop.

As a result, timelines for developing and deploying our products and services may be longer than we currently

anticipate, and we may encounter delays due to, among other things, technical challenges, including those resulting

from the nascent state of certain of our products and services, the unavailability or immaturity of key technologies,

supply chain constraints, energy shocks, including related price volatility, labor availability, permitting and

regulatory approvals, or the need to redesign or reengineer key components. In addition, the costs associated with

developing and deploying our products and services and related platforms, infrastructure and strategic initiatives at

scale may exceed our current estimates, including due to inflationary pressures, energy prices, unforeseen

engineering complexities, the cost of developing or licensing technologies that are not yet commercially available,

competitive dynamics, changes in scope, or the need for additional capital expenditures, contingency reserves or

working capital.

If we are unable to successfully execute our growth strategy on the anticipated timeline or within our expected cost

parameters, our business, financial condition and results of operations could be materially adversely affected. Delays

or cost overruns could also impact our ability to achieve projected returns, meet contractual commitments, access

additional financing on acceptable terms, or maintain investor confidence. Moreover, even if we successfully deploy

our growth strategy, including Starship, Terafab, orbital AI, and the creation of the lunar economy, they may not

perform as expected at scale, which could result in operational inefficiencies, increased costs, reduced revenues, or

declines in our stock price.

We have experienced, and will likely continue to experience, launch delays and failures that could have a

material adverse effect on our business, financial condition, results of operations, and future prospects.

Launch vehicle underperformance, propulsion anomalies, structural failures, software errors, or other malfunctions

could result in launch delays or partial or total mission failures, including the loss of satellites or payloads. The

occurrence of mission failures or other significant operational disruptions could also expose us to litigation as well

as increased scrutiny from regulatory authorities, lead to the imposition of additional compliance requirements, and

adversely affect our brand and reputation, and our ability to obtain future licenses, permits, or government contracts.

We do not typically obtain insurance coverage for our satellites, payloads, or launch vehicles, and as a result we bear

the full financial cost of any such losses. Repeated anomalies or high visibility mission failures could also negatively

affect our brand, reputation, ability to win new business, and our customers’ ability to procure launch and in-orbit

insurance at competitive rates (to the extent we decide to pursue it). Such repeated anomalies or mission failures

could also result in, regulators delaying, conditioning or denying approvals, waivers or licenses required for future

launches or reentries, which could reduce our launch cadence and delay the deployment of our satellites and other

services. In the past, certain of our launch vehicles have experienced partial or total mission failures, including

anomalies that resulted in the loss of payloads and damage to launch vehicles. In certain circumstances, such

mission failures could result in, debris from our launch vehicles causing significant damage to persons or property

on the ground as well as environmental damage. There can be no assurance that similar or other failures will not

occur with future launches. In addition, satellites may be deployed into incorrect or suboptimal orbits due to vehicle

performance issues, separation events, or guidance, navigation and control errors. Incorrect orbital placement can

33

materially reduce a satellite’s operational life, impair performance, increase fuel consumption, or render the satellite

unusable.

Our satellites, launch vehicles, and other space-related technologies operate, and in the case of orbital AI

compute, will operate, in the harsh and unpredictable environment of space, exposing them to a wide and unique

range of space-related risks that could cause them to malfunction or fail, and any such malfunction or failure

could adversely affect our business, financial condition, results of operations, and future prospects.

Operating in space subjects our satellites, launch vehicles, spacecraft, and related systems to extreme and highly

variable conditions that can adversely affect performance, reduce useful life, or result in total mission failure. Space

is inherently hostile. Hardware must withstand: significant vibration and acoustic loads during launch; wide-ranging

thermal cycles; radiation from solar and cosmic sources; micrometeoroids and orbital debris; and other

environmental hazards, each of which testing cannot fully replicate. In particular, we have not, and no one else has,

previously operated or attempted to operate orbital AI compute, and the conditions of space on such AI

infrastructure have not been tested. Once deployed, orbital AI compute infrastructure will not be readily accessible,

and as a result, will not be easily repaired or upgraded, such that any component failures could result in permanent

capacity loss, accelerated depreciation, decommissioning or need for replacement of the infrastructure.

In addition, space weather events, such as geomagnetic storms, solar flares, and other forms of radiation activity,

have in the past disrupted and could in the future disrupt satellite propulsion, power systems, and communications

equipment, potentially leading to reduced performance or permanent damage. Although we incorporate certain

radiation-hardened components, shielding, and redundancy into our systems, these measures may not be sufficient to

prevent material adverse impacts in all scenarios. Failures or performance degradation resulting from these risks

could delay deployments, reduce available capacity, increase operating costs, require significant capital expenditures

to replace affected assets, or interrupt or degrade services provided to customers. Furthermore, the useful life of our

satellites is inherently shorter than that of the information technology systems and infrastructure they host. As a

result, we must periodically launch replacement satellites as existing satellites reach the end of their useful lives and

are decommissioned, which may truncate the effective lifespan of those underlying information technology systems

and infrastructure. Any such events could adversely affect our reputation, compliance with applicable laws and

regulations, business, financial condition, results of operations, and future prospects.

The continued proliferation of satellite constellations in Low-Earth Orbit, as well as the risk of collisions with

space debris or other spacecraft, could limit or impair our launch flexibility and satellite deployment, which could

adversely affect our business, financial condition, results of operations, and future prospects.

The continued proliferation of Low-Earth Orbit constellations can increase the risk of collisions with space debris or

other spacecraft if operators fail to adhere to responsible space safety, debris mitigation, or coordination practices.

Our growth strategy depends, in part, on continuing to launch additional satellites into Low-Earth Orbit. As the

number of satellites and other objects in Low-Earth Orbit continues to grow, the probability of accidental collisions,

fragmentation events, or other in-orbit incidents increases, which could result in the loss or degradation of our

satellites, increased costs for collision avoidance maneuvers, or the need to replace or reposition assets on an

accelerated schedule. Not all satellite operators or other space actors adhere to the same rigorous space safety, debris

mitigation, or coordination practices that we adhere to, which may increase the likelihood of congestion,

conjunctions, or other operational risks outside of our control and, in extreme cases, could contribute to

fragmentation events or cascading debris effects that further increase collision risks in Low-Earth Orbit.

In addition, some domestic and international authorities have applied heightened regulatory scrutiny as interest in

utilizing Low-Earth Orbit for satellite operations has increased. Debris mitigation regulations may emerge if

congestion increases. Failure to meet debris requirements could result in monetary penalties or loss of licensing

authority, which would adversely affect our satellite constellation deployment and expansion plans, and future

regulatory actions could impose more restrictive operational, deployment, or debris mitigation requirements that

could limit our ability to launch or operate satellites in Low-Earth Orbit. In addition, there is a burgeoning effort to

further regulate Low-Earth Orbit, MEO, and GSO and establish liability regimes for operators, including regimes

similar to those under the Comprehensive Environmental Response, Compensation and Liability Act, which imposes

strict liability for environmental contamination or remediation costs, as well as growing concern over the potential

34

environmental effects of emissions and other byproducts from rocket launches in Earth’s upper atmosphere.

Additional regulation in this area could adversely impact our business, financial condition, results of operations, and

future prospects.

Furthermore, any damage to our satellites or impairment of their functionality resulting from collisions with space

debris or other spacecraft could materially and adversely affect our ability to deliver reliable services to our

customers, harm our reputation, and expose us to potential contractual liabilities or insurance claims. The growing

challenges associated with space debris management may require us to invest in additional technologies or processes

to safeguard our assets and maintain compliance with evolving regulatory frameworks, which could have a material

adverse effect on our business, financial condition, results of operations, and future prospects.

Interruptions in the operation of critical satellite network, ground station, launch, manufacturing, or spacecraft

or data center infrastructure could result in significant downtime, operational delays or loss of service, each of

which could have a material adverse effect on our business, financial condition, results of operations, and future

prospects.

Our ability to provide reliable services across our Space, Connectivity, and AI business segments depends on the

uninterrupted operation of our critical infrastructure, including but not limited to satellite and communications

networks, ground stations, launch facilities, and data centers. An interruption or failure affecting any aspect of this

infrastructure, whether due to equipment malfunctions, power outages, disruptions in, or unauthorized access to, our

computer systems (such as software or hardware failures, or cyberattacks), natural disasters (such as earthquakes,

floods, fires, or severe weather events), terrorism, war, sabotage, pandemics, epidemics, or other unforeseen

circumstances, could result in significant downtime, operational delays, or complete loss of service. Any such attack

could destroy or disable a significant number of our satellites and, depending on its scale, could trigger a cascading

collision event that renders our licensed orbits, and potentially other orbits, unusable for an extended period.

Similarly, the use of our satellites to enable communications access in conflict zones may expose us to retaliation

from foreign governments and non-state actors. Such an event could have a material adverse effect on our business,

financial condition, results of operations, and future prospects. These events may disrupt power, damage facilities,

interrupt service despite contingency plans or compromise our ability to deliver services to customers as promised,

hinder our ability to meet regulatory or contractual requirements, and erode trust among our customers, partners,

regulators and stakeholders. In particular, an interruption or failure affecting our critical infrastructure could result in

outages of service to our Starlink Subscribers. Any such outage could erode the trust of existing and potential

Starlink Subscribers in our service, which could result in the loss of existing or potential subscribers. In addition, the

complexity and interdependence of our engineering, manufacturing, assembly and terrestrial, space transportation,

and infrastructure systems mean that a disruption in one component can have cascading effects throughout our

operations. For example, an outage at a data center or ground station could impact command and control functions,

mission planning, or real-time telemetry, while interruptions at launch facilities could cause postponements or

cancellations of scheduled launches.

Adverse global macroeconomic and geopolitical conditions may negatively affect our business, financial

condition, results of operations and future prospects.

Adverse global or regional economic and geopolitical conditions could reduce demand for certain of our products

and services, which may negatively affect our business, financial condition, results of operations and future

prospects. Economic downturns, inflation, higher interest rates, tighter credit conditions, reduced consumer

spending, lower business or government investment, or geopolitical developments may negatively affect demand for

our offerings. Reduced consumer or enterprise spending for each of our Starlink connectivity services or our AI-

related offerings would limit our ability to grow our business, which may slow the pace at which we deploy satellites

and expand our constellation or adversely affect the utilization of our launch capabilities.

35

Manufacturing, testing and launching rockets, satellites, and spacecraft, including our efforts to reuse rockets

and spacecraft, involve inherent risks that could result in human injury or death, property damage and

environmental damage or other adverse environmental impacts due to accidents or equipment failures. Any such

events could result in substantial losses, including reputational harm and legal liability, which could have a

material adverse effect on our business.

The manufacturing, testing, launching, and recovery of our rockets, satellites, and spacecraft are complex activities

that are conducted under challenging conditions and involve a high degree of risk. Our reusable vehicles will reenter

Earth’s atmosphere and fly over populated land for extended periods, which carries inherent risks to populations in

the event of failure, such as structural breakup, loss of control, or debris dispersal. Although we implement extensive

safety protocols and operational safeguards designed to protect personnel and the public, these protocols and

safeguards may not in all circumstances prevent exposure of our personnel and potentially members of the public to

hazards such as explosions, structural failures or debris dispersal. A manufacturing defect, testing anomaly, launch

failure, recovery incident, or similar event involving injury to humans, any human fatalities, property damage, or

environmental damage or other adverse environmental impacts could result in substantial losses, including

reputational harm and legal liability, which could have a material adverse effect on our business.

Although we are focused on the vertical integration of our businesses, we depend on third parties to manufacture

and supply certain key components necessary for the provision of our launch, connectivity, and AI services, and

any supply shortages or disruptions or failures in their performance could have a material adverse effect on our

business, financial condition, results of operations, and future prospects.

Disruptions in the supply chain for essential raw materials or components, challenges in the supplier qualification

process, or increases in the prices of inputs could materially and adversely affect our business, financial condition,

results of operations, and future prospects. Despite our supply chain being largely vertically integrated, our reliance

on third-party manufacturers and suppliers for key components introduces risks related to supply chain continuity,

quality assurance, and vendor performance. We depend on both domestic and international suppliers for certain

specialized materials, components, and services that are essential to the production and operation of our launch

vehicles, spacecraft, satellites, user terminals (including Starlink consumer terminals), AI segment and related

infrastructure. Any failure or delay by these partners to deliver components in the required quantities, within

specifications, or on schedule has in the past and may in the future adversely affect our production schedules,

operational reliability, and our ability to meet contractual obligations. In addition, disruptions in the supply chain

due to shortages, quality issues, natural disasters, geopolitical events, labor disputes, pandemics, epidemics, tariffs or

trade restrictions, criminal activity (including terrorism, sabotage or cyberattacks) or other factors outside our

control could result in significant delays, increased costs, or an inability to deliver products and services to

customers in a timely and cost-effective manner. The process of qualifying new suppliers or transitioning to

alternative vendors can be time-consuming and may not be successful, further increasing our exposure to supply

chain interruptions. Furthermore, our limited pool of qualified vendors for certain critical products or services

exposes us to increased pricing pressures and quality risks. In particular, certain materials and products that are key

inputs in our Space, Connectivity, and AI segments are available from a limited number of suppliers, including sole

or limited-source suppliers, and our direct chip suppliers are dependent on a concentrated group of advanced

semiconductor fabrication facilities. For additional information regarding supply chain risk relating to our AI

processors, please see “Our ability to scale our AI products relies on our terrestrial and orbital AI compute

infrastructure, which depends on the availability of power, AI processors, and other critical components,

telecommunications services, and any shortages or disruptions thereof would materially adversely affect our

business, financial condition, results of operations, and future prospects.” The inability of these suppliers to deliver

necessary components of the products in a timely manner and at prices, quality levels, and volumes acceptable to us,

or interruptions in supply of materials or products on which these suppliers rely, could have an adverse effect on our

ability to meet customer demands and contractual obligations, execute on our growth strategy, or manage our

expenses or timelines as expected, which could adversely impact our business, financial condition, results of

operations, and future prospects.

36

Our ability to scale our AI products relies on our terrestrial and orbital AI compute infrastructure, which

depends on the availability of power, AI processors, and other critical components, telecommunications services,

and any shortages or disruptions thereof would materially adversely affect our business, financial condition,

results of operations, and future prospects.

Our ability to scale our data center infrastructure, which supports our AI segment, is increasingly constrained by the

availability of power at economically feasible prices, long lead times, availability of materials, and changing

regulatory requirements. For example, energy supply is constrained globally due to the significant increase in

demand for, and limited availability of, energy to power AI compute. Securing this capacity can involve entering

into complex, long-lead-time arrangements or proceeding with alternative sources of power generation. We

currently rely significantly on natural gas and gas turbine technology to power our data center operations. As such,

our ability to scale our infrastructure depends in part on our continued access to natural gas supply at economically

feasible prices, the availability of gas turbines and related equipment, and the maintenance of a regulatory

environment that permits and supports the use of natural gas for large-scale power generation. Our AI products also

rely on GPUs and other processors, servers, network equipment and other critical components sourced from third-

party suppliers for use in our data centers. Manufacturing and supply of servers and network equipment for our

technical infrastructure, particularly for GPUs and other specialized components, is limited to a small number of

qualified suppliers. We do not have any long-term or other material contractual arrangements with our direct chip

suppliers, instead procuring all of our GPUs on a purchase-order basis. Our direct chip suppliers are dependent on a

concentrated group of advanced semiconductor fabrication facilities, or “fabs.” Any disruption to our upstream

supply chain, including fab capacity constraints, manufacturing issues, shortages of raw materials such as silicon

wafers or rare earth elements, geopolitical tensions affecting fab operations, or natural disasters impacting key

fabrication regions, could limit our chip suppliers’ ability to fulfill our orders, which could have a material adverse

effect on our business, financial condition, and results of operations. Our ability to achieve orbital AI at scale

depends on our ability to access a sufficient number of AI chips, significantly more than are currently available to

us. While we expect to construct Terafab to address such supply constraints, Terafab may not be successful, in

which case we may not have other sources of sufficient AI chips to meet our orbital AI compute demands. While

Terafab is intended to expand our internal chip manufacturing capabilities and alleviate potential future AI chip

shortages at SpaceX, particularly as we pursue orbital AI at scale, we expect to continue sourcing a significant

portion of our compute hardware from third-party suppliers, and there can be no assurance that we will be able to

achieve our objectives with respect to Terafab within the expected timeframes, or at all. While we have a framework

agreement with Tesla, neither Tesla nor Intel are obligated to remain a part of the project, and we may not enter into

any such definitive agreements. Our AI segment also relies on services from third-party telecommunications

providers, including connectivity to the cloud, and internet bandwidth suppliers to provide uninterrupted and error-

free services through their networks. We may be unable to obtain AI processors or other necessary components or

telecommunications services at prices or volumes that are acceptable to us or in a timely manner. Our suppliers and

telecommunications and internet service providers also serve other customers, including certain of our competitors,

and such suppliers or providers may prioritize capacity for such other customers, increase prices on short notice,

require onerous prepayments, or reduce or delay deliveries to us. Any failure by our suppliers and service providers

to meet our cost, quality, volume, or delivery requirements, or any shortage or disruption in the supply of chips,

telecommunications services or other components required for our AI segment, could result in service disruption or

outages, delay critical data center or network infrastructure upgrades or expansions, impair our ability to train our AI

models and meet customer demand for our AI segment products and materially adversely affect our business,

financial condition, results of operations and future prospects.

We also rely on third-party cloud compute providers for a portion of the compute used for the X platform and may

from time to time rely on third-party data center providers, which exposes us to several risks that are beyond our

direct control, including vulnerability to outages, performance issues, and cyberattacks. We have non-cancellable,

multi-year capacity commitments to cloud compute providers, requiring payment regardless of usage. A termination

or lapse in service from third-party cloud compute and data center providers could expose us to service interruptions,

significant delays, and additional expenses to re-architect products for a different provider. Additionally, in the event

of nonperformance by us or our providers, or an industry downturn, we may incur liabilities, have excess capacity

that we cannot easily redeploy, and fail to receive payments from our counterparties or customers.

37

We face intense competition in the markets in which we operate, and while we have historically outperformed

certain competitors in our Space and Connectivity segments, we may not continue to do so, which could adversely

affect our business, financial condition, results of operations, and future prospects.

The markets in which we operate are rapidly evolving and intensely competitive, and we face competition from a

range of established and emerging companies, including large, well-capitalized technology companies and aerospace

firms, including foreign competitors. Some competitors are investing significant capital to develop and deploy

satellite constellations and related infrastructure that compete directly with our offerings, and companies based in

China and other jurisdictions may benefit from government support, favorable regulatory environments, or strategic

national prioritization.

Some of our current and potential competitors, particularly in our AI segment, have greater financial, technical,

manufacturing, or other resources than we do, and may devote more resources to the development and

commercialization of competing products and services. Competitors may adopt more aggressive pricing, secure

more favorable supplier or distribution arrangements, bundle services, form strategic alliances or otherwise take

actions that enhance their competitive position in ways that could adversely affect our business. In certain markets,

regulatory or geopolitical factors may result in preferential treatment for domestic competitors or otherwise limit our

ability to compete effectively.

Competition continues to intensify as new technologies are developed and new entrants emerge. While we have

historically outperformed certain competitors in aspects of our business, such as our Space and Connectivity

segments, there can be no assurance that we will maintain this position.

We depend on our ability to recruit and retain employees who have advanced engineering and technical skills,

and intense competition for such employees may increase costs and affect our ability to meet development and

production timelines.

We depend on our ability to recruit and retain employees who have advanced engineering and technical skills and, in

some cases, employees with the necessary national security clearances to perform under our government contracts or

win new business. These employees are in great demand and are likely to remain a limited resource in the

foreseeable future. The current tight labor market has adversely impacted our ability to recruit qualified personnel,

including engineers, particularly with respect to our AI segment. Increased restrictions on the import or retention of

foreign labor may also increase demand for engineering personnel and adversely impact our ability to hire and retain

qualified personnel. Continued turnover may impact employee morale and create other challenges as we attempt to

scale our AI business. In addition, significant amounts of time and resources are required to train technical and other

personnel, and we have in the past lost and may in the future lose new employees to our competitors or other

companies before we realize the benefit of our investment in recruiting and training them. Our ability to recruit and

retain qualified employees depends on a number of things, including our ability to pay market compensation,

provide opportunities for advancement, and secure visa sponsorships and work permits for qualified international

candidates. If we are unable to recruit and retain a sufficient number of these employees, then our ability to maintain

our competitiveness and grow our business could be negatively affected. In addition, because of the highly technical

nature of our products and services, the loss of any significant number of our existing engineering personnel could

have a material adverse effect on our business, financial condition, results of operations, and future prospects. A

significant portion of the talent pool for advanced engineering and technical roles is international, and changes in

immigration laws or policies in the jurisdictions in which we operate could limit our ability to hire and retain such

candidates and intensify competition for talent.

From time to time, we are involved in litigation, investigations, and other regulatory proceedings which could be

costly, time-consuming, and divert management attention, materially adversely affecting our business.

From time to time, we have been and may in the future become involved in various legal proceedings relating to a

variety of matters, including intellectual property, commercial, regulatory, product liability, employment, personal

injury, class action, employee or contractor health and safety, environmental, whistleblower, securities and other

litigation and claims, and governmental and other regulatory investigations and proceedings, including tax

examinations. Additionally, our share price may be volatile and, in the past, companies that have experienced

38

volatility in the market price of their stock have been subject to securities litigation, including class action litigation.

Such matters could be costly, time-consuming, and divert management’s attention from executing our strategic

initiatives and operating our business. The industries in which we operate have historically experienced significant

litigation and regulatory scrutiny, and with our public profile, expanding operations and the novel nature of some of

our offerings, including our AI solutions, we may face an increased risk of such actions. Litigation and regulatory

proceedings are inherently unpredictable. Any adverse judgments, settlements, or regulatory penalties could result in

substantial financial costs, reputational harm, and operational disruptions. Certain of our hardware products are new

and relatively unproven. If a product defect were to arise, especially one leading to product liability claims, the

resulting warranty and damage claims, together with any associated harm to our reputation, could have a material

adverse effect on our business, financial condition, results of operations, and future prospects. Even if we prevail in

these matters, the defense and resolution of litigation and regulatory proceedings may require significant resources

and management attention, which could materially and adversely affect our business, financial condition, results of

operations, and future prospects. Additionally, the mere initiation of litigation or government inquiries, regardless of

the outcome, could negatively impact investor confidence and our stock price. As we continue to innovate and

pursue new commercial and government contracts, expand our product offerings, and enter new markets, the

likelihood of facing legal and regulatory challenges may increase, further exposing us to these risks. Please refer to

“Business—Legal Proceedings” and Note 17, Commitments and Contingencies, in our audited consolidated

financial statements and Note 16, Commitments and Contingencies in our unaudited consolidated financial

statements include elsewhere in this prospectus.

Any significant disruption in, or unauthorized access to, our computer and data systems or those of third parties

that we utilize in our operations could result in a loss or degradation of service, loss of trust in us and harm to

our business.

An operational disruption in, or unauthorized access to, our computer and data systems or those of third parties that

we utilize in our operations could compromise sensitive (including classified or otherwise government-controlled),

proprietary, confidential, or personal information, impede operations, and result in financial losses, legal liabilities,

reputational harm, and erosion of our competitive position in launch services, space-based internet, and mobile

phone services. Our business depends on the continuous and secure operation of our information technology systems

and infrastructure, including those that support our launch operations, manufacturing facilities, Starlink services,

government services, employee databases, and mission-critical communications. Our systems and infrastructure may

also be subject to cyberattacks, including sophisticated hacking attempts by nation-states, state-sponsored actors,

cybercriminals, or other malicious third parties, which could result in unauthorized access to, disruption of, or

degradation of our satellite systems, ground infrastructure, or data networks. Such disruptions or unauthorized

access, which may result from a wide variety of incidents or activities, including inadvertent compromises arising

from process, coding or human errors, cyberattacks, data breaches, exploitation of known or unknown software or

hardware vulnerabilities, malware, ransomware, credential harvesting, computer viruses, social engineering (such as

phishing), denial of service attacks, software or hardware failure, or other malicious or disruptive incidents or

activities—whether perpetrated by external actors, including nation-states, state-sponsored organizations, or

cybercriminal groups, insiders, or other threat actors, any of whom may see their efforts enhanced by the use of AI

—could lead to the theft, destruction, or unauthorized disclosure of sensitive (including classified or otherwise

government-controlled), proprietary, confidential or personal information, including technical data, customer or

partner information, and intellectual property, particularly because some of our products and services involve the

collection, storage, and processing of such data and information. Our development and deployment of AI models,

internal and third-party AI tools, and other AI applications expose us to increased and novel risks and

vulnerabilities, including prompt injection, hallucinations, errors, and other issues related to AI agents, as well as the

risk of compromise of valuable intellectual property including source code, model weights, and other assets. Certain

internal and external threat actors, such as nation-states, state-sponsored organizations, organized threat networks

and corporate espionage actors, among others have and will continue to sustain malicious activities for extended

periods and deploy significant resources to attempt, and in some cases succeed, at causing significant disruptions in,

or unauthorized access to, our computer systems or those of third parties that we utilize in our operations. Such

incidents have in the past and may in the future also disrupt or degrade our ability to design, produce, launch, or

manage our products and services, resulting in operational delays, violations of applicable data privacy and

39

cybersecurity laws and regulations, disruptions in, or unauthorized access to, our customers’ computer systems,

increased costs, loss of revenue, loss of trust, litigation or regulatory penalties.

As the scale, frequency, sophistication, or intensity of cyber and data privacy threats continue to evolve, and as our

reliance on interconnected systems and third-party vendors grows, we remain exposed to vulnerabilities despite our

efforts to implement security measures, monitoring, and incident response protocols. There can be no assurance that

our cybersecurity risk management processes, including our policies, procedures, and controls, will be effective in

promptly or effectively detecting, containing, or remediating cybersecurity attacks. Any significant security and data

breach or system failure could materially and adversely affect our business, financial condition, results of operations,

and future prospects, and could result in loss of trust among customers, regulators, government agencies, and

partners. Furthermore, our efforts to investigate, mitigate, contain, and remediate the harm caused by a significant

disruption in, or unauthorized access to, our computer and data systems or those of third parties that we utilize in our

operations may be costly and time-consuming and may not be successful, and we may make errors or fail to take

necessary actions. Remediation efforts, litigation, regulatory investigations, and compliance obligations (including

obligations to notify appropriate regulators and affected parties) arising from such incidents could require substantial

management attention and resources, and we rely on our own funds to cover such losses or liabilities. In addition,

rapid changes to U.S. and international cybersecurity and privacy laws and regulations have expanded regulatory

regimes and compliance requirements, and regulators continue to undertake enforcement actions in these areas. We

expect the regulatory environment to grow more complicated, which may increase our operational and compliance

expenditures, as well as those of our suppliers. Moreover, some third parties we utilize in our operations may receive

or store information provided by us or by our customers. If these third parties fail to adopt or adhere to adequate data

privacy and security practices, or their systems or networks are breached in the manner described above, our data or

our customers’ data may be improperly accessed, used, or disclosed to unauthorized recipients, which could result in

financial losses, legal liabilities, reputational harm, and additional compliance obligations. We do not control the

privacy and cybersecurity measures put in place by such third parties, and any contractual protections with such

third parties, such as obligations to indemnify us, if any, may be ineffective or otherwise inadequate.

The development and maintenance of the technologies and infrastructure necessary to support our current and

future operations will require significant capital expenditures, and if we are unable to generate sufficient cash

flow from operations or obtain additional financing on acceptable terms, our business, financial condition,

results of operations, and future prospects could be materially and adversely affected.

Our business requires substantial capital expenditures to design, develop, expand, and maintain our technologies and

infrastructure to support our operations. For example, we have incurred significant capital expenditures and expect

to increase our capital expenditures substantially in the future in connection with the design, development, and

deployment of our satellite constellations, launch vehicles, ground stations, manufacturing facilities, and programs,

including Terafab, AI compute infrastructure, data centers, and other supporting infrastructure. These expenditures

include, but are not limited to, costs associated with research and development, construction and expansion of

production capabilities, acquisition of property and equipment, and ongoing maintenance and upgrades to ensure

reliability and competitiveness. In particular, the development, testing, and deployment of Starship in accordance

with our anticipated schedule, as well as our pursuit of orbital AI, other space-related services, and lunar and

interplanetary missions, will require the investment of significant additional capital resources. In addition, we have

made and intend to continue to make substantial capital expenditures to support the growth of our AI products,

including costs related to obtaining third-party GPUs, manufacturing our own GPUs, and constructing, leasing,

maintaining, enhancing, and expanding our data centers. We may choose to increase or accelerate the pace of any of

these investments at any time, which could result in periods of reduced profitability or increased losses as we

prioritize long-term growth over near-term financial performance. Many of the products and services that are

important for our growth prospects are novel and untested, and therefore our estimates of capital expenditures may

prove to be inaccurate.

If we raise additional capital through further issuances of equity or convertible debt securities, our shareholders

could suffer significant dilution and any new equity securities we issue could have rights, preferences, and privileges

superior to those of holders of our Class A common stock. The agreements governing our indebtedness contain

various restrictive covenants and any additional debt financing secured by us in the future could involve restrictive

covenants relating to our capital-raising activities and other financial and operational matters, which could limit our

40

operational flexibility and make it more difficult for us to obtain additional capital and to pursue business

opportunities. Our ability to access the capital markets or secure other sources of financing may be adversely

affected by factors beyond our control, including fluctuations in market conditions, changes in investor sentiment,

increases in interest rates, or adverse events affecting the broader industry or economy.

Our substantial level of indebtedness could materially adversely affect our financial condition.

We have significant indebtedness that could materially adversely affect our business by increasing our vulnerability

to general adverse economic and industry conditions; requiring us to dedicate a substantial portion of our cash flow

from operations to payments on our indebtedness, thereby reducing the availability of our cash flow to fund

operations, our growth strategy, product development and strategic initiatives; limiting our flexibility in planning

for, or reacting to, changes in our business and the industry in which we operate; and exposing us to the risk of

increased interest rates as our borrowings are, and may in the future be, at variable interest rates. As of March 31,

2026, we had total principal indebtedness outstanding of $29,132 million. Our substantial indebtedness may also

adversely affect our credit ratings or outlook, which may increase our cost of capital, limit our access to financing,

and impair our ability to obtain additional financing on acceptable terms, or at all. The occurrence of any one of

these events could have a material adverse effect on our business, results of operations, and financial condition, and

ability to satisfy our obligations under the agreements governing our indebtedness. If we fail to comply with the

terms of our debt agreements, our lenders could declare a default and accelerate our repayment obligations, which

could materially and adversely affect our business, financial condition, results of operations, and future prospects.

Our future revenue and operating results depend upon our ability to develop new technologies and respond to

changes in customer demands and industry standards in highly competitive markets, and if we are unable to do

so, our business, financial condition, results of operations, and future prospects may be materially and adversely

affected.

Our future revenue growth and operating results are highly dependent on our ability to design, develop and

successfully commercialize new and innovative technologies, products, and services on a timely and cost-effective

basis. The markets in which we operate are characterized by rapid and disruptive technological change, evolving

industry standards, the emergence of new and well-funded competitors, frequent new product and service

introductions, changing customer demands and regulatory changes. In addition, we may expand into new markets,

which may lead to similar or additional challenges that we cannot foresee and may require novel innovations to

navigate or overcome. As a result, we may from time to time rapidly adjust, modify or change our strategic

priorities, capital allocation, product or service focus or operational initiatives across our business in response to

these other changes or new markets. In particular, the AI industry is nascent, highly competitive, capital intensive

and rapidly changing. There are a number of companies today that develop or may develop products or services that

compete with our AI segment, and new competitors may emerge over time. Some of our current or potential

competitors in the AI market are large technology companies that have significant financial, technical and marketing

resources, and in some cases greater access to data, and others are smaller specialized companies that possess

specialized expertise and may have greater flexibility than we do. We also have a limited number of customers for

our AI products when compared to certain of our competitors. Current and potential competitors have established, or

may in the future establish, cooperative relationships among themselves or with third parties to increase the ability

of their AI technologies to address the needs of current and prospective users of our AI products. Furthermore,

current or prospective users may decide to develop competing products for particular use cases or to establish

strategic relationships with our competitors for such use cases. Current and potential competitors and bad actors,

may also attempt to reverse engineer or otherwise replicate our AI technology, including through model extraction

or distillation techniques. Increased competition with our AI products could result in price reductions, revenue

shortfalls, loss of customers and loss of market share, which may harm our business, financial condition results of

operations and future prospects.

In our Connectivity segment, including Starlink broadband and Starlink Mobile, we face competition from terrestrial

fixed network providers, mobile network operators, and other satellite providers, and our services may be less

competitive in certain markets, including dense urban areas where terrestrial fiber and wireless networks may offer

higher capacity, lower cost, or more consistent performance. In addition, our Starlink Mobile offering operates in a

highly competitive and evolving market, and may be affected by the pace of technological development, spectrum

41

availability, and the success of our partnerships with mobile carriers. In addition, the X platform faces intense

competition from social media, messaging and media companies and traditional media outlets, such as television,

radio and print, for advertising budgets. Advertisers generally do not have long-term commitments to the X platform

and may reduce or discontinue their advertising spending for a variety of reasons outside our control. We are

expending resources to improve the X platform and improve its attractiveness to users and advertisers. While we

have introduced new user interface enhancements, algorithm updates, and other product features, improvements to

the X platform, introducing new products and services on the X platform and other initiatives may be costly and

difficult to implement, and we cannot be sure that they will be positively received by users, content creators, or

advertisers, or provide positive returns on our investment. Losing users who migrate to other platforms may

negatively impact our potential subscription or advertising revenue. Additionally, if users do not continue to

contribute content and otherwise engage with the X platform, we are unable to provide users with valuable and

timely content, or if content that is considered to be problematic or offensive is made available on the X platform,

the size of the X platform’s user base and their engagement may decline, leading to a decline in monetizable usage

and the loss of potential subscription revenue from such users, and the X platform may experience brand or

reputational harm. A decline in users on the X platform, or the volume or quality of their content on the X platform,

could also impact the ongoing development of our AI product, which in part utilizes data and user-generated content

from the X platform. We plan to publicly launch the Money product on the X platform (the “Money Product”);

however, we are competing against large, established companies with significantly greater resources and market

presence than us. If we are unable to anticipate technological trends, respond to technological advancements or

changing customer demands, or successfully develop and commercialize new or enhanced offerings, we may be

unable to establish or maintain a meaningful market position and our business, financial condition, results of

operations, and future prospects could be materially and adversely affected.

The estimates of future market opportunity and forecasts of market growth, and our ability to capture such

markets, included in this prospectus may prove to be inaccurate.

Our estimates for the total addressable market for our Space, Connectivity and AI businesses, as well as estimates

regarding the growth of AI and its impacts, contained elsewhere in this prospectus are based on a number of internal

and third-party estimates. For example, our estimates of market opportunity for our Space, Connectivity and AI

businesses rely in part on third-party data and a number of internal assumptions. With respect to our Space segment,

these estimates rely in part on estimates published by Novaspace regarding the size of the global market for space-

enabled solutions, including spacecraft manufacturing, launch services and related activities. Our connectivity

market estimates are based in part on estimates of the number of households, businesses, aircraft and maritime

vessels globally derived from third-party sources, together with assumptions regarding ARPU and monthly service

revenue derived from third-party industry data and our internal expectations regarding pricing, adoption rates and

service penetration across different geographic regions and economic environments. Our AI market estimates are

based in part on projections of global data center compute demand from third-party sources, including estimates

published by RAND Corporation, together with internal assumptions regarding the portion of global compute

capacity that may be utilized for AI workloads and other operational assumptions such as power usage, utilization

rates and pricing.

These estimates require us to make numerous assumptions and judgments regarding factors that are inherently

uncertain and subject to change, including the pace of technological development, future demand for launch,

connectivity and AI services, the rate of adoption of satellite connectivity and AI technologies, the availability and

cost of power and computing hardware, the evolution of regulatory frameworks, and broader macroeconomic

conditions.

While we believe our assumptions and the data underlying our estimates are reasonable, these assumptions and

estimates may not be correct and the conditions supporting our assumptions or estimates may change at any time,

thereby reducing the predictive accuracy of these underlying factors. As a result, our estimates of the total

addressable market for our services, as well as the expected growth rate for the total addressable market for our

services, may prove to be inaccurate.

42

Many of our initiatives, including those to develop orbital AI compute at scale, manufacture AI chips at scale,

establish a lunar economy, develop human augmentation systems, and transport humans and cargo to the Moon

and Mars, involve significant technical complexity, unproven technologies, or technologies that do not exist or

may require significant advancement, and such initiatives may not achieve commercial viability.

Our initiatives to develop orbital AI compute at scale, establish a lunar economy, develop human augmentation

systems, and transport humans and cargo to the Moon and Mars are in early stages of conception, design and

development and have not yet been proven at commercial scale, or at all, and may ultimately be unsuccessful. In

particular, the timeline for these initiatives, and the launch cadence required to achieve them may be difficult or

impossible to determine. These efforts require substantial and ongoing investments of financial, technical, and

human resources over extended time horizons, including, but not limited to, research and development, testing,

infrastructure, regulatory approvals, and mission execution. The technologies, systems, and operational capabilities

required for each of these initiatives involve significant technical complexity and are subject to design, engineering,

and performance risks, many of which may only become apparent as development and testing progress. Many of

these technologies, systems and operational capabilities are novel and untested, and we expect to incur significant

capital expenditures over a period of years before our AI products and services and other strategic initiatives,

including AI compute infrastructure and in-orbit, lunar, and interplanetary industrialization efforts, become

profitable, which may never occur. In addition, in-orbit refueling of Starship is essential to our lunar, Mars, asteroid

mining, and other deep space ambitions beyond geostationary Earth orbit. In-orbit refueling is complex, and we have

not yet demonstrated or attempted it. We may not be able to develop, commercialize, scale, or successfully

implement these or other strategic initiatives on the timelines we currently anticipate, or at all. Furthermore, the

viability of orbital AI compute depends in part on the cost advantages of solar energy relative to existing terrestrial

energy sources. To the extent that breakthrough developments in terrestrial energy access, such as advances in

nuclear energy, significantly reduce energy costs or alleviate infrastructure constraints, the viability of our orbital AI

compute infrastructure may be materially diminished. Even if our orbital AI compute infrastructure proves to be

commercially viable, a material slowdown in the growth of AI applications and related compute demand could result

in existing terrestrial data centers sufficiently meeting such demand, thereby reducing the need for our orbital AI

compute infrastructure. As a result, we may be required to devote financial, technical, human or other resources in

excess of our current expectations, and there can be no assurance that these investments will generate adequate

revenue, which could adversely affect our business, financial condition, results of operations, and future prospects.

Several of our anticipated market opportunities, including certain AI, orbital, lunar, and interplanetary

transportation and industrial activities, are still emerging and evolving or do not currently exist, and such

markets may not develop as we expect, or at all.

A portion of our anticipated market opportunities is associated with industries described in the section entitled

“Business—Future Markets.” Certain of these industries, such as space tourism, human augmentation, and cargo

transport to the Moon, are still emerging. Others, including in-orbit manufacturing, passenger transport to the Moon,

an established human presence or gateway hub on the Moon, passenger and cargo transport to Mars, energy

production on the Moon or Mars, manufacturing capabilities on the Moon or Mars, and asteroid mining do not exist

today. Any estimate we make regarding the size or timing of our anticipated market opportunities is inherently

uncertain and necessarily involves significant assumptions about future customer demand, adoption, technological

development, regulatory conditions and the emergence of a broader commercial market that does not currently exist.

While we believe these industries will develop over time, the manner in which they emerge, including the timing of

commercialization, the scale and pace of adoption, and the applicable technical, regulatory, geopolitical and

economic frameworks may differ materially from our current expectations. If these industries do not develop,

develop on slower timelines, at smaller scales, or under different economic or regulatory conditions than we

anticipate, this could require us to modify, delay, or abandon certain of our business plans, or cause such plans not to

develop at all, which could materially and adversely affect our business, financial condition, results of operations,

and future prospects.

43

The global nature of our business poses risks with respect to unstable, malicious or arbitrary legal regimes and

authorities.

We, particularly through Starlink, maintain global operations. As a result, we may face risks that our operations will

be subject to unstable, capricious, or malicious legal regimes and authorities. The increasing militarization of space

and the potential development of space-based warfare capabilities may expose our assets and operations to

heightened geopolitical and security risks, including the risk that foreign governments or other actors could target

our satellites or related infrastructure. Certain foreign governments have publicly discussed the potential use of anti-

satellite weapons against the Starlink constellation. These and other actions by foreign governments, whether

through military, regulatory or other means, may adversely affect our operations and assets. Even if we attempt to

comply with known local laws, our assets (both physical, intangible and financial) may be subject to seizure or other

expropriation. There is no guarantee that we will be able to maintain operations in any jurisdiction, and, if our assets

or properties are subject to seizure or other expropriation, there can be no assurances that we will be able to recover

our assets or properties. Any such legal or other governmental action could have an adverse effect on us. For

example, in August 2024, Starlink received an order from Brazil’s Supreme Court that froze Starlink’s Brazilian

financial assets and prevented Starlink from conducting financial transactions in Brazil (the “Brazil Asset Seizure”).

The action taken by the Brazilian Supreme Court arose out of purported violations of Brazilian law by X, which at

the time was not owned by us and was only affiliated with Mr. Musk. It is possible that we may be subject to actions

like the Brazil Asset Seizure in the future (whether in Brazil or another country) and, regardless of whether any such

action is consistent with local and international law, we may never recover assets seized in any similar action.

Additionally, actions that we take to minimize the impact of actions such as the Brazil Asset Seizure to our

customers, for example, by continuing to provide service without charge or otherwise altering payment processes

and methods to permit customers to maintain service, may have a material impact on our financial performance. As

evidenced by the Brazil Asset Seizure, we may be subject to adverse actions from governmental actors on the basis

of assumptions, facts or events that are not directly related to our operations and instead relate to the actions of our

directors, officers, or shareholders or operations of businesses that are affiliated with them.

Our services are subject to risks related to supplying services to the U.S. government.

Supplying services to the U.S. government subjects us to unique risks, including compliance with complex

regulations, vulnerability to changes in government priorities or funding levels, and exposure to contractual disputes

or audits. In 2025, approximately one-fifth of our revenue was attributable to agencies within the U.S. federal

government. As a contractor to various U.S. government agencies, we are subject to extensive federal procurement

regulations, including the Federal Acquisition Regulation (FAR) and Defense Federal Acquisition Regulation

Supplement (DFARS), as well as other rules governing cost accounting, cybersecurity, ethics, and national security.

These regulations impose stringent requirements on our operations, business practices, and reporting, and

noncompliance could result in civil or criminal penalties, suspension or debarment from government contracting, or

loss of existing or future business. These requirements, although customary in U.S. government contracts, increase

our performance and compliance costs. These costs might increase in the future. For those reasons and in order to

achieve our orbital compute goals, we may prioritize our own launch payloads over additional U.S. government

contracts or third-party customers. This prioritization of launch capacity may limit revenue growth in our Space

segment, and impact our relationship with regulators, and could invite litigation from customers or competitors. In

addition, government contracts are susceptible to unilateral termination, reduction in scope, or delays at the

government’s convenience, which may occur due to shifting budgetary priorities, changes in defense or space

policy, or the reallocation of funding to other programs. The termination or reduction of funding for a government

program could result in a loss of anticipated future revenue attributable to that program. The actual receipt of

revenue on awards may never occur or may change because a program schedule could change or the program could

be canceled, or a contract could be reduced, modified, or terminated early. In addition, in certain circumstances,

governments or other customers may be reluctant to rely on our satellite connectivity or defense-related services if

they believe the availability of such services could be restricted or suspended based on geopolitical considerations,

conflicts, sanctions, or other policy determinations, which could adversely affect our ability to win or retain

contracts. In addition, our significant business relationships with U.S. defense and government agencies may cause

us to be perceived as closely aligned with the U.S. government or military. This perception could discourage certain

consumers, enterprises, or foreign governments from purchasing our products and services which could adversely

44

affect our sales in the United States and internationally. We and our facilities could also be targeted by foreign

adversaries and non-state actors due to such perception. Government customers may also subject our contracts to

rigorous audits and investigations, which can result in disputes regarding contract performance, cost allowability, or

compliance with applicable laws and regulations. Adverse audit findings or contractual disputes could lead to

repayments, financial penalties, or restrictions on our ability to compete for future contracts.

Certain of our government contracts also require that we maintain facility security clearances and that certain of our

employees obtain and maintain personnel security clearances. Obtaining and maintaining these clearances involves a

lengthy and uncertain process and depends on factors outside of our control, and we may experience delays in

receiving required clearances or be unable to hire or retain a sufficient number of employees with the necessary

clearances to perform under certain contracts. If we are unable to obtain or maintain required facility or personnel

security clearances, we may be unable to bid on, win, or perform certain classified programs, and existing contracts

could be terminated or not renewed, which could materially and adversely affect our business, financial condition,

results of operations, and future prospects.

Further, our business is subject to economic sanctions and trade embargo laws, various import regulations, including

tariffs, and stringent U.S. import and export control laws. Any failure by us to comply with any of the foregoing

could result in our debarment from government contracts, limitations on our ability to enter into contracts with the

U.S. government, civil or criminal penalties, fines, investigations, more onerous compliance requirements, or loss of

export privileges.

We derive significant revenue from U.S. government contracts that are subject to competitive bidding, funding

approvals and other government budgetary processes, which factors could adversely affect our business, financial

condition, results of operations, and future prospects.

We derive significant revenue from U.S. government contracts that were awarded through a competitive bidding

process. Competitive bidding presents a number of risks, including: the need to bid on programs in advance of the

completion of their design, which may result in unforeseen technological difficulties and cost overruns; the

substantial cost and managerial time and effort that must be spent to prepare bids and proposals for contracts that

may not be awarded to us; the need to estimate accurately the resources and cost structure that will be required to

service any contract we are awarded; and the expense and delay that may arise if interested parties or our

competitors protest or challenge contract awards made to us pursuant to competitive bidding, and the risk that any

such protest or challenge could result in the delay of our contract performance, the distraction of management, the

resubmission of bids on modified specifications, or in termination, reduction or modification of the awarded

contract.

Our business with governmental entities is subject to changes in policies, priorities, regulations, mandates, and

funding levels, any of which could materially impact our operations and financial results. U.S. government program

funding is subject to Congressional appropriations on a fiscal year basis even though contract performance may take

more than one year. As a result, at the outset of a major program, the contract is usually incrementally funded and

additional funds are normally committed to the contract only as Congress makes appropriations in future fiscal

years. U.S. government contracts may also be undefinitized at the time of the start of performance. Under

undefinitized contract actions, the U.S. government has the ability to unilaterally definitize contracts and, absent a

successful appeal of such action, the unilateral definitization of the contract would obligate us to perform under

terms and conditions imposed by the U.S. government. Such unilaterally imposed contract terms could include less

favorable pricing or terms and conditions more burdensome than those negotiated in other circumstances. U.S.

government contracts typically involve long lead times for design and development and are subject to significant

changes in contract scheduling.

Additionally, the U.S. government’s budget deficit and the national debt, as well as any inability of the U.S.

government to complete its budget process for any government fiscal year and consequently having to shut down or

operate on funding levels equivalent to its prior fiscal year pursuant to a “continuing resolution,” could have a

material and adverse impact on our business, financial condition, results of operations, and future prospects.

Moreover, if we fail to establish and maintain important relationships with U.S. government agencies, our ability to

successfully maintain and develop new business could be materially and adversely affected. The current political

45

environment in the United States is highly polarized, and shifts in the composition of the U.S. Congress or changes

in the presidential administration can result in significant changes in government spending priorities, regulatory

posture, and the allocation of contracts and resources across industries and programs. Our relationships with U.S.

government agencies and the favorability of the regulatory and procurement environment in which we operate may

be affected by which political party controls the presidency or one or both chambers of the U.S. Congress. As a

result, there can be no assurance that current government relationships, contracts, or levels of funding will be

maintained, and any significant adverse developments could have a material and adverse impact on our growth and

competitive position.

In addition, our Space segment revenue is primarily derived from fixed-price contracts, under which we agree to

deliver specified products or services at a predetermined price regardless of the actual costs incurred. As a result, if

we experience cost overruns on these contracts, including from factors outside our control, we are required to absorb

the excess costs, which may reduce profitability or result in losses, strain cash flows, and impact our ability to invest

in future growth. Any unanticipated increases in labor, material, or other direct or indirect costs—including those

arising from inflation, supply chain disruptions, design changes, regulatory requirements, or unforeseen technical

challenges—must be borne by us. When these overruns occur, our margins on affected contracts may be

significantly reduced or eliminated, which could adversely affect our business, financial condition, results of

operations, and future prospects. Additionally, absorbing excess costs may limit our ability to allocate resources to

other strategic initiatives, delay investment in research and development, or constrain our capacity to pursue new

business opportunities. In addition, we sometimes receive advanced payments and billings in excess of the amount

of revenue we recognize, which we record as deferred revenue. As a result, our cash flows may be subject to

fluctuation across periods in a manner that may be unrelated to our underlying performance.

Our ability to expand our Starlink consumer and enterprise connectivity services depends on our ability to

increase market awareness and acceptance of connectivity through Starlink, and any failure to do so could

materially and adversely affect our business, financial condition, results of operations and future prospects.

Our ability to expand our Starlink consumer and enterprise connectivity services depends on our ability to increase

market awareness and acceptance of connectivity through Starlink. There can be no assurance that our efforts to

increase awareness will be successful. In particular, such efforts may not be successful if we are unable to offer

Starlink services at competitive prices. Additionally, constraints in the distribution of user terminals could delay

service activations, increase costs, or otherwise limit our ability to scale such services as anticipated. Consumer

acceptance may also be hindered by the presence of well-established terrestrial broadband alternatives, as well as

lingering perceptions regarding service reliability, latency, and the complexity of satellite-based internet compared

to traditional fixed-line solutions.

The expansion of our satellite-to-mobile connectivity services depends substantially on our ability to secure and

maintain partnerships with mobile network operators and on the adoption of necessary hardware and software

modifications by device manufacturers, and any failure to do so could materially and adversely affect our

business, financial condition, results of operations and future prospects.

The expansion of our global satellite-to-mobile connectivity offerings depends substantially on our ability to enter

into and maintain successful partnerships with telecommunications carriers and spectrum licensees globally, and to

obtain country-specific authorizations to offer such connectivity using satellite spectrum bands for which we have

international coordination rights. In the United States, we expect to be able to provide 5G-like connectivity to a

meaningful portion of existing unmodified devices through our Starlink Mobile Gen2 service utilizing our V2

Mobile satellites, either by operating on spectrum leased to us by MNO partners or by utilizing our own domestic

spectrum holdings. However, achieving full 5G NR-NTN compliance and optimal performance would likely require

handset manufacturers to implement hardware and software modifications, primarily to the radio-frequency front

end, in future devices. The spectrum frequencies in the FCC licenses to be acquired from EchoStar are standardized

for terrestrial 5G mobile broadband (3GPP bands n66 and n70). But the 5G NR-NTN bands for these same

frequencies, such as n252 and n256, are not currently supported by RF front-end hardware for the provision of 5G-

like service in any commercially available mobile devices. We do not have direct contractual arrangements with

handset manufacturers; instead, we expect MNO partners, as major purchasers of mobile devices, to encourage or

46

drive such adoption. There can be no assurance that these modifications will be adopted on our preferred timeline, or

at all.

Internationally, we face similar constraints until handset manufacturers implement hardware and software

modifications to support the international spectrum authorizations to be obtained from EchoStar. As a result, our

near-term international service strategy depends on our ability to establish MNO spectrum partnerships on a market-

by-market basis, which does not require device hardware modifications but is subject to the successful negotiation

and execution of commercial agreements in each jurisdiction. Until device manufacturers incorporate support for our

international spectrum bands into future handsets, we will be unable to offer 5G-like direct-to-consumer service on

our own international spectrum.

The provision of our satellite-to-mobile services also requires regulatory approvals from the FCC and foreign

regulatory authorities. Our Gen1 service, utilizing our existing constellation of V1 Mobile satellites, is fully licensed

in the United States but requires additional country-by-country approvals to operate internationally. We have signed

MNO partnerships for our Gen1 service in over 30 countries. These partnerships represent commercial agreements

with carriers but do not, by themselves, provide the regulatory approvals necessary to offer service. In addition to at

least one MNO partnership, we have obtained required approvals to offer commercial Gen1 service in the United

States, Canada, the United Kingdom, Japan, and Australia, as well as in several additional countries.

Our Gen2 service, which will utilize 2 GHz S-band spectrum and a new satellite constellation, requires a license

transfer, a constellation license, and spectrum usage approvals in each country in which we seek to operate. For the

United States, we have received the relevant license transfer approval from the FCC, and we expect to receive the

remaining necessary regulatory authorizations in the second or third quarter of 2026. While these authorizations

would be sufficient from a United States regulatory perspective, we still require our V2 Mobile satellites to be in

orbit and must complete the acquisition of the relevant spectrum from EchoStar before we can commence our

planned commercial Gen2 service in the United States. Internationally, we have filed applications in nearly every

country in which we intend to operate our Gen2 service, and approvals have been granted in a limited number of

these jurisdictions to date. Each jurisdiction presents its own regulatory process and timeline, and we cannot predict

when or whether approvals will be granted in any given market. Subject to regulatory approvals, we are receiving

from EchoStar certain assets and authorizations that provide very senior ITU priority for international frequency

coordination for our V2 Mobile constellation. Until such approvals are obtained, we also signed a coordination

agreement with EchoStar to obtain the protection of its senior ITU priority rights until the authorizations transfer.

However, some countries have signaled through public consultations or other actions that they are considering

ignoring or diminishing ITU priority as a mechanism to decide which operators are licensed to operate in their

country. Several countries and regions have open inquiries that invite input on whether factors other thanITU

priority (such as whether the operator originates from the country) should govern the issuance of spectrum licenses,

and we cannot be certain the outcome of these proceedings.Delays or failures to obtain necessary approvals could

materially delay the deployment and commercialization of our Gen2 service. The failure to enter into or successfully

maintain such partnerships, or the failure of device manufacturers to adopt the necessary hardware modifications, or

the failure to obtain required regulatory approvals, could materially and adversely affect our business, financial

condition, results of operations, and future prospects.

If the recommendations, forecasts, content, analyses or other output that our AI technologies, including Grok,

assist in producing are or are alleged to be deficient, inaccurate, harmful, illegal, or used for an improper

purpose, we could continue to be subjected to claims and investigations, and we could be subjected to legal

liability and brand, reputational, or competitive harm.

AI technologies, the models, algorithms, prompts and datasets on which they rely, and the recommendations,

forecasts, analyses or other output that such AI technologies assist in producing, may be flawed, insufficient, of poor

quality, rely upon incorrect, inaccurate, harmful or illegal data, reflect unwanted forms of bias, hallucinate,

misrepresent, mislead or contain other errors or inadequacies, any of which may not be easily detectable. Although

we devote significant resources to develop, test, and maintain our AI technologies, we may not be able to identify or

resolve all AI-related issues, deficiencies, and failures before they arise. AI technologies have been known to

produce mischaracterized or “hallucinatory” inferences or outputs, and certain of our AI products, such as Grok,

have been alleged to be susceptible to “data poisoning” in the past. We may not have insight into, or control over,

47

the practices of third parties who may utilize our AI technologies. As such, third parties have in the past used, and

may in the future use, such AI technologies for improper purposes, including through the dissemination of illegal,

inaccurate, defamatory or harmful content, intellectual property infringement or misappropriation, furthering bias or

discrimination, cybersecurity attacks, including spear phishing and social engineering attacks, data privacy

violations, other societal harms, including activities that threaten people’s safety, financial security, or mental well-

being on- or offline, or to develop competing technologies. Inappropriate or controversial data practices by data

scientists, engineers, and end users of AI technologies, including our AI segment’s systems, could impair the

acceptance of AI technologies generally, including our AI products. If the recommendations, forecasts, content, or

analyses that our AI technologies assist in producing are or are alleged to be deficient, inaccurate, offensive, illegal,

or otherwise harmful, we could be subjected to claims and investigations, and we could be subjected to legal liability

and brand, reputational or competitive harm. We have in the past been, and may in the future be, subject to

regulatory investigations and litigation related to such claims regarding our recommendations, forecasts, content or

analyses. Also please refer to “—Our AI products, X platform, and Starlink services are subject to complex and

evolving U.S. and foreign laws and regulations regarding privacy, cybersecurity, data use, data combination, data

protection, content, AI, competition, youth protection, safety, consumer protection and notification, advertising, e-

commerce, sanctions, export controls,and other matters. Many of these laws and regulations are subject to change

and uncertain interpretation, and we could be required to make changes to our products and business practices, and

be exposed to monetary penalties, increased cost of operations, declines in user growth or engagement, or loss of

customers, or other harm to our AI products, X platform, and Starlink services.” In addition, if we do not have

sufficient rights to use the models, algorithms, prompts and datasets on which our AI technologies rely, or the

recommendations, forecasts, content, analyses or other output that our AI technologies assist in producing, we could

also incur liability through the violation of applicable laws and regulations, third-party intellectual property, privacy

or other rights, or contracts to which we are a party. Furthermore, failure to properly disclose the use of consumer-

facing AI technologies may result in consumer protection or regulatory enforcement activity. Use of AI

technologies, including our AI products, may result in disruptions in, or unauthorized access to, users’ computer

systems, which could also lead to the unauthorized disclosure of sensitive (including classified), proprietary,

confidential or personal information, new potential cyberattack methods for third parties or an increase in the

frequency, sophistication or intensity of cyberattacks. Moreover, if AI technologies are perceived to be significantly

disruptive to society, it could lead to governmental or regulatory restrictions or prohibitions on their use, societal

concerns or unrest, or both, any of which could materially and adversely affect our ability to develop, deploy, or

commercialize AI technologies and execute our business strategy. Our implementation of AI technologies, including

through our AI segment’s systems, could result in legal liability, regulatory action, operational disruption, brand,

reputational or competitive harm, or other adverse impacts.

Environmental laws, regulations, litigation, liabilities and proceedings may adversely affect our operations,

including our launch operations, manufacturing activities, fuel storage and handling operations, launch facilities

and ground infrastructure, and data center operations and expansion plans.

Our operations, including our launch operations, manufacturing activities, fuel storage and handling operations,

launch facilities and ground infrastructure, and data center operations and expansion plans are subject to a variety of

state and federal environmental laws and regulations governing matters such as air emissions, wastewater discharges

and the discharge, treatment, storage, disposal and remediation of hazardous substances and wastes, including the

Comprehensive Environmental Response, Compensation and Liability Act, the Resource Conservation and

Recovery Act, the Clean Air Act, the Clean Water Act and permitting requirements of federal, state and local

environmental authorities. Liability under these laws imposes strict liability for environmental contamination or

remediation costs. Changing regulatory requirements for permits and approvals relating to operational infrastructure,

including energy generation assets (e.g., renewables, generators or grid connections), manufacturing facilities,

launch facilities, fuel storage and handling facilities, and data centers may cause delays, higher costs or denials, and

a failure to comply with these requirements may result in fines, shutdowns or competitive harm. In addition,

growing scrutiny of data centers’ overall ecological footprint could lead to community opposition, fines or mandates

for changing existing practices. We are or may become subject to environmental lawsuits and proceedings, and

various parties have threatened or brought lawsuits that allege we are unlawfully operating natural gas-fired turbines

without required permits at facilities in Southaven, Mississippi. While we have obtained such permits, the outcome

of these legal actions is uncertain. Injunctive relief or the rescission of issued permits would prevent our ability to

48

utilize power generation sources that are required for the operation of these data centers and would adversely affect

our AI business. We cannot predict with certainty how future legislative or regulatory developments will affect our

business, but compliance with new or modified environmental requirements could require us to incur significant

unanticipated expenditures that could adversely affect our financial condition, results of operations, speed of

deployment and cash flows.

In addition, our launch facilities and related operations are subject to environmental permitting, land use, wetlands,

coastal management and other environmental review requirements, such as the National Environmental Policy Act

or related federal and state laws, that may give rise to litigation, regulatory enforcement actions or permitting

disputes. Environmental groups, regulatory authorities or other stakeholders may challenge our launch activities,

launch cadence, construction or expansion of facilities, fuel storage or handling practices, or other operational

activities under federal, state or local environmental laws. Such actions may seek injunctive relief, civil penalties or

additional environmental review and mitigation measures, any of which could delay launches, restrict operations,

increase compliance costs or otherwise adversely affect our business, financial condition, results of operations and

future prospects.

We may face substantial potential liability and operational disruptions if we violate the intellectual property rights

or other rights of third parties, and if we fail to adequately protect, maintain, defend or enforce our intellectual

property and other similar rights, we could lose an important competitive advantage, in each case which could

have a material adverse effect on our business, financial condition, results of operations, customer trust and

future prospects.

Our success and ability to compete also depends in part on our ability to operate without infringing,

misappropriating or otherwise violating the intellectual property rights of third parties. Companies in the AI and

technology industries own large numbers of patents, copyrights, trademarks, and trade secrets, and frequently enter

into litigation based on allegations of infringement, misappropriation, or other violations of intellectual property or

other rights, including in novel areas such as those relating to AI training and AI outputs. Plaintiffs have in the past

and may in the future file infringement or other litigation or administrative or adversarial actions relating to the

training or development of our AI models. We cannot guarantee that the operation of our business does not and will

not infringe or violate the rights of third parties, and we may be unaware of the intellectual property rights that

others may claim cover some or all of our products or services. Moreover, we may not have the freedom to operate

unimpeded by the patent or other rights of others. Third parties may have dominating, blocking or other patents or

other rights relevant to our technology, of which we are not aware.

Intellectual property and related laws are constantly evolving, can be highly uncertain and involve complex legal

and factual questions for which important principles remain unresolved. For example, in the United States and in

many foreign jurisdictions, policies regarding the breadth of claims allowed in patents and scope of protections for

content can be inconsistent. We cannot predict future changes in the interpretation of patent, intellectual property

and other related laws or changes to patent, intellectual property and other related laws that might be enacted into

law by U.S. and foreign legislative bodies.

We rely on statutory safe harbors, including those set forth in the Digital Millennium Copyright Act and Section 230

of the Communications Decency Act in the United States and the Digital Services Act in the EU, to protect against

liability for various activities, including linking, caching, ranking, recommending and hosting. Legislation or court

rulings affecting these safe harbors may harm us and may impose significant operational challenges. There are

legislative proposals and pending litigation in the United States, EU, and around the world that could diminish or

eliminate safe harbor protection for websites and online platforms.

If we violate, or are alleged to have violated, the intellectual property rights of third parties, including patents,

copyrights, trademarks, trade secrets, or other intellectual property rights and related rights, we may be subject to

costly and time-consuming litigation, substantial financial penalties, and reputational harm, any of which could

materially disrupt our operations, product development and strategic initiatives. As we continue to develop new or

update existing technologies, products, and services, there is a risk that third parties may allege that our operations

or offerings infringe upon their intellectual property rights. For example, we are currently a defendant in litigation

alleging copyright infringement relating to the claimed use of copyrighted works to train our AI models. Other

49

plaintiffs may file infringement or other litigation relating to the training or development of our AI models. In

addition, we are currently subject to, and in the future may be subject to claims from various “non-practicing

entities” or other companies that own patents and other intellectual property rights that often attempt to aggressively

assert their rights in order to extract value from technology companies by threatening costly litigation or that have

minimal operations or relevant product revenue and against whom our patents may provide little or no deterrence or

protection. We are and may in the future be subject to additional copyright litigation or other litigation, including

litigation relating to allegations that we have trained or developed our AI models on copyrighted works in a manner

that infringes on copyrights, or in a manner that otherwise violates the intellectual property or other rights of third

parties, or that our models produce outputs in a manner that infringes on copyrights or other intellectual property or

other rights. Moreover, the impact of AI on intellectual property ownership and licensing rights, including

copyrights, has not been fully addressed by U.S. or international courts or other federal, state or international laws or

regulations (or by courts, laws or regulations in foreign jurisdictions), and our use of AI models may reduce our

ability to protect our own intellectual property. In addition, former employers of our current, former, or future

employees may assert claims that such employees have improperly disclosed to us confidential or proprietary

information of these former employers. Any such claims or allegations, whether or not they have merit, could result

in costly litigation, substantial damages, injunctions against the use of certain technologies, or the need to obtain

licenses on unfavorable terms. In addition, certain of our contracts with customers, suppliers, and partners contain

indemnification provisions that could require us to defend against infringement or other claims and pay damages or

settlements, thereby increasing our financial exposure. The outcome of intellectual property litigation is inherently

uncertain, and adverse judgments could materially and adversely affect our business, financial condition, results of

operations, and future prospects. If we are unable to obtain necessary licenses, non-infringing substitute

technologies, or otherwise mitigate these risks, we may be forced to discontinue certain products or services, delay

or curtail research and development activities, or limit our expansion into new markets.

Additionally, failure to adequately protect, maintain, defend, or enforce our intellectual property—including patents,

copyrights, trademarks, trade secrets, and proprietary technologies—may lead to loss of competitive advantage,

weakened market position, and financial harm from unauthorized use or infringement. We rely and expect to

continue to rely upon a combination of patents, trademarks, trade secrets, copyrights, confidentiality procedures,

contractual commitments and other legal rights to establish and protect our intellectual property. However, the steps

we take to protect our intellectual property and other rights may be inadequate due to various circumstances. We

may be unable or choose not to pursue or maintain certain types of intellectual property protection or registration for

our intellectual property in the United States or foreign jurisdictions, and the measures we do take may not prevent

our competitors or other third parties from independently developing products, services, and technology similar to or

duplicative of our products and services. We will not be able to protect our intellectual property if we are unable to

enforce our rights or if we do not detect unauthorized use of our intellectual property. In addition, our patents or

other intellectual property rights may be challenged, invalidated, circumvented or rendered unenforceable, and

pending and future trademark and patent applications may not be approved. While it is our policy to enter into

confidentiality agreements with our employees, contractors and other third parties to limit and control access to and

disclosure of our trade secrets, intellectual property and confidential information, we may fail to enter into such

agreements with all relevant entities and any such agreements may be breached, or this intellectual property may

otherwise be disclosed or become known to our competitors, including through hacking, theft, or other

misappropriation, including by employees, which could cause us to lose any competitive advantage resulting from

these trade secrets, intellectual property and proprietary information. Accordingly, we cannot guarantee that the

steps we have taken to protect our intellectual property will be adequate to prevent infringement of our rights or

misappropriation of our technology, trade secrets or know-how.

Additionally, to protect our intellectual property rights, we may be required to spend significant resources to

monitor, defend, enforce and protect these rights. Monitoring unauthorized uses of our intellectual property is

difficult and costly. We may not be able to detect unauthorized use of, or take appropriate steps to enforce, our

intellectual property rights. Litigation may be necessary in the future to enforce our intellectual property rights and

to protect our trade secrets, and any such litigation may be costly and time consuming, result in the diversion of time

and attention of our management team, and may not be successful or could result in the impairment or loss of

portions of our intellectual property. Furthermore, our efforts to enforce our intellectual property rights may be met

with defenses, counterclaims and countersuits attacking the validity and enforceability of our intellectual property

50

rights. Despite our efforts, we may not be able to prevent unauthorized use, copy, reverse engineering,

misappropriation of our technology or intellectual property rights to create technology that compete with ours, or

independent development of similar technologies. Insufficient protection could force us into costly and uncertain

litigation or enforcement actions, allowing competitors to launch rival products and eroding our revenue and

profitability.

Acquisitions, divestitures, or other strategic transactions we pursue may not achieve the anticipated benefits,

synergies or strategic objectives.

We may not achieve the anticipated benefits, synergies, or strategic objectives of any acquisition, divestiture, or

other strategic transaction in a timely manner, or at all, including those we expect from the recent acquisition of xAI,

the acquisition of spectrum assets and licenses from EchoStar in connection with our Starlink Mobile initiatives, our

collaboration on Terafab with Tesla, Intel or any future partners, project and our recent collaboration with Cursor

and any potential acquisition of Cursor, if consummated. Acquisitions, divestitures, or other strategic transactions

may present unforeseen liabilities or disruptions to our operations, which could adversely impact our business,

financial condition, results of operations, and future prospects. We may assume unexpected obligations or incur

costs associated with acquired businesses, including litigation, regulatory compliance, environmental liabilities, or

contractual disputes, which could result in material losses or divert management focus from ongoing operations.

Integrating acquired businesses, partnerships, or joint ventures may present significant challenges, including

aligning operations, systems, and cultures, which could result in inefficiencies, increased costs, or failure to realize

anticipated benefits. The process of integration is often complex and time-consuming, and we may encounter

unforeseen difficulties in harmonizing business practices, integrating technologies and IT systems, retaining key

personnel, or reconciling differences in corporate cultures and management philosophies. In addition, the integration

of acquired entities or new partners exposes us to disruptions in, or unauthorized access to, our computer systems

and data or may divert management attention and resources from our core operations, potentially impacting our

ability to execute on other strategic initiatives or maintain existing customer relationships. We may also face

challenges in achieving expected synergies, cost savings, or strategic objectives within anticipated timeframes, or at

all, which could adversely affect our business, financial condition, results of operations, and future prospects. If we

are unable to successfully integrate acquisitions, partnerships, or joint ventures, or if the anticipated benefits of these

transactions do not materialize as expected, we could experience operational disruptions, loss of key personnel or

customers, increased costs, and diminished competitive position. Any failure to effectively integrate acquired

businesses, partnerships, or joint ventures could materially and adversely affect our business, financial condition,

results of operations, and future prospects.

Similarly, divestitures could result in the loss of revenue, disruption of customer or partner relationships, or

challenges in separating assets and personnel. There can be no assurance that we will be able to identify,

consummate, or integrate future acquisitions, divestitures, or other strategic transactions on favorable terms, or at all,

and any such activities may heighten our exposure to operational, financial, and regulatory risks unique to our

industry.

We have experienced, and will likely continue to experience, development and manufacturing delays and damage

or destruction during pre-launch operations, any of which could have a material adverse effect on our business,

financial condition, results of operations, and future prospects.

The development, manufacturing, and operation of launch vehicles and satellites are complex and capital-intensive

activities that are subject to numerous risks. Our launch vehicles, satellites, and related systems have in the past

experienced and may in the future experience delays, damage or destruction during design and manufacturing,

including delays in fabrication, assembly, inspection, testing, and component qualification. These issues may arise

from engineering challenges, supplier performance problems, quality control shortcomings, unexpected design

modifications, or disruptions in our manufacturing facilities. Any of these factors may delay development or

production schedules, increase costs, or result in hardware that must be reworked or replaced.

Our operations also involve significant risks during pre-launch preparation. Launch vehicles and satellites can be

damaged or destroyed during transport, fueling, integration, or ground testing. Furthermore, the early retirement or

51

inoperability of satellites or related infrastructure may require us to accelerate depreciation or recognize impairment

charges, thereby adversely affecting our business, financial condition, results of operations, and future prospects.

Even minor anomalies may require extensive troubleshooting or repairs, resulting in launch delays, increased

mission costs, or the loss of flight hardware. Because launch operations require coordination across multiple systems

—including propulsion, avionics, ground infrastructure, and third-party range providers—issues in any one area can

lead to postponements or mission cancellations.

Our ability to continue and expand launch and satellite operations depends upon our ability to obtain new and

leverage existing U.S. export control and sanctions authorizations, and any significant changes to the geopolitical

landscape or U.S. government regulatory approach to licensing could materially and adversely impact our

international business operations by compromising existing licenses or limiting our ability to engage in

commercial dealings in or involving geopolitically sensitive countries.

The launch and satellite operations are subject to stringent export control and economic and trade sanctions laws,

including the U.S. International Traffic in Arms Regulations (“ITAR”), the Export Administration Regulations, and

sanctions administered and enforced by the U.S. Treasury Department’s Office of Foreign Assets Control

(“OFAC”). Under U.S. export control laws, we are required to obtain export authorizations from the Departments of

Commerce or State to export or share any controlled goods, technology, or software with foreign persons, including

foreign person employees, or to foreign destinations. The availability of such authorizations may be impacted by

significant changes to the geopolitical landscape. The U.S. government may revise export control regulations,

restrict exports to new or additional locations, or otherwise change its approach to licensing in ways that, while

outside of our control, materially impact our international supply chain, existing export licenses, and business

operations. For example, under the ITAR, we are required to determine the proper licensing jurisdiction and

classification of products, software and technology; and obtain licenses or other forms of U.S. government

authorizations to engage in certain activities related to and that support our business operations. The authorization

requirements include the need to get permission to release controlled technology to foreign person employees and

other foreign persons.

In addition, we are required to obtain OFAC authorization in certain situations, including to provide connectivity

services or engage in other business operations in certain global markets that may be subject to economic sanctions

or trade embargoes. While we have been successful in obtaining such authorizations in the past, there can be no

assurances that authorizations or licenses will be available in the future. In addition, significant changes to the

geopolitical landscape, such as the outbreak of armed conflict, could result in the imposition of new or expanded

economic or trade sanctions that may impact or prevent our ability to provide services or otherwise operate in certain

markets. Failures by us to comply with import, export control, or sanctions laws and regulations could result in civil

or criminal penalties, fines, investigations, more onerous compliance requirements, loss of export privileges,

debarment from government contracts, or limitations on our ability to enter into contracts with the U.S. government.

Other regulators, such as the EU or UK, may also impose restrictions on our ability to operate in geopolitically

sensitive countries or territories.

Our use of open source technology could impose limitations on our ability to commercialize our space-based

internet and mobile phone services, AI products, and X platform, or otherwise negatively affect our business.

We use open source technology in some of our software, including in our Starlink products and services, and in our

AI segment’s and X platform’s software and products, and we expect to continue to use open source technology in

the future. Open source technology is licensed by its authors or other third parties under open source licenses, which

in some instances may subject us to certain unfavorable conditions. For example, certain open source licenses may

give rise to requirements to disclose or license our proprietary source code or make available any derivative works

or modifications of the open source code on unfavorable terms or at no cost. Although we monitor and have

implemented policies relating to our use of open source technology to avoid subjecting our products and services to

conditions we do not intend, we cannot guarantee such efforts will be successful and we may face allegations from

others alleging ownership of, or seeking to enforce the terms of, an open source license, including by demanding

release of the open source software, derivative works or modifications, or our proprietary source code that was

developed using such technology, or demanding access to our software free of charge or on other unfavorable terms.

These allegations could also result in litigation. Additionally, our AI products are trained on data sets that may

52

include open source software, and it is possible that certain outputs of our AI products may be subject to open source

license restrictions or obligations. The terms of many open source licenses are ambiguous and have not been

interpreted by United States or foreign courts. There is a risk that these licenses could be construed in a way that

could impose unanticipated conditions or restrictions on our ability to commercialize our AI segment’s products. In

such an event, we may be required to seek licenses from third parties to continue commercially offering our AI

segment’s products, to make our proprietary code generally available in source code form, to re-engineer our AI

segment’s products or to discontinue the sale of our AI segment’s products or such other products if re-engineering

could not be accomplished on a timely basis, any of which could adversely affect our business, financial condition,

results of operations, and future prospects.

In addition, the use of open source technology may entail greater technical and legal risks than those associated with

the use of third-party commercial software as open source licensors generally do not provide support, warranties,

controls on origin of the software, indemnification or other contractual protections regarding infringement claims or

the quality of the code, including the existence of security vulnerabilities. Many of the risks associated with usage of

open source technology, such as the lack of warranties or assurance of title, cannot be eliminated and could, if not

properly addressed, negatively affect our business. To the extent that our technologies and other business operations

depend upon the successful and secure operation of the open source technology we use, any undetected errors or

defects in this open source software could prevent the deployment or impair the functionality of our software, delay

the introduction of new technological capabilities, result in a failure of our technologies, and injure our brand and

reputation. For example, undetected errors or defects in open source software could render it vulnerable to breaches

or security attacks and make our AI segment’s products more vulnerable to data breaches or security attacks. Any of

the foregoing would have a material adverse effect on our business, financial condition, results of operations and

future prospects.

Payment, banking, and other financial service-related activities may subject us to additional regulatory

requirements, regulatory actions, and other risks that could be costly and difficult to comply with or that could

harm our business.

We plan to publicly launch the Money Product, which will offer payment, banking and other financial services

functionalities, including enabling our users to purchase tangible, virtual, and digital goods from merchants and send

money to other users, among other activities. These activities will subject us to a variety of laws and regulations in

the United States, Europe, and elsewhere globally, including those governing anti-money laundering and counter-

terrorism financing, money transmission, stored value, gift cards and other prepaid access instruments, electronic

funds transfer, virtual currency, consumer protection, charitable fundraising, global and local economic sanctions,

and import and export restrictions. In addition, we could become subject to new consumer protection laws and

regulations that may be adopted or amended, including those related to payment, banking, and other financial

services activities as well as sharing, collection, and use of payment, banking, and other financial services-related

data. Depending on how the Money Product evolves, we may also be subject to other laws and regulations including

those governing gambling, cryptocurrencies, brokerage, banking, credit, and lending. In some jurisdictions, the

application or interpretation of these laws and regulations is not clear. We have received certain payments licenses in

the United States and other jurisdictions for our anticipated regulated payments-related products and activities.

These licenses increase flexibility in how our use of payments may evolve, help mitigate regulatory uncertainty, and

will generally require us to demonstrate compliance with many domestic and foreign laws in relation to our licensed

payments products and activities. Our efforts to comply with these laws and regulations may still not guarantee

compliance. In the event that we are found to be in violation of any such legal or regulatory requirements, we may

be subject to monetary fines or other penalties such as a cease and desist order, or we may be required to make

product changes, any of which could have a material and adverse effect on our business, financial condition, results

of operations and future prospects.

In addition, we will be subject to a variety of additional risks as a result of payment, banking, and other financial

services transactions, including: increased costs and other resources to address errors in transactions or customer

disputes; potential fraudulent or otherwise illegal activity by users, developers, employees, or third parties;

restrictions on the investment of consumer funds used to transact payments; and additional disclosure and reporting

requirements. We plan to publicly launch the Money Product and may in the future undertake additional payment,

53

banking, and other financial services initiatives, which may subject us to many of the foregoing risks and additional

licensing requirements.

Our efforts to support the creation of permanent installations on the Moon and Mars depend on the successful

development and deployment of next-generation capabilities.

Activities related to the industrialization and development of the Moon and Mars require the successful development

and deployment of next-generation capabilities such as fully reusable launch vehicles, including Starship, in-space

refueling and propellant storage, in space communications systems, and other capabilities required for operations

beyond Earth’s orbit. These systems involve significant technological, engineering, and operational challenges,

including the need to develop habitable transportation and surface environments, and perform complex in-orbit

operations. Solving these challenges will require developing solutions that are novel or untested and will require

substantial capital investment. If these efforts take longer than anticipated, or if technical, operational, or engineering

challenges arise in connection with these efforts, our goals with respect to the Moon and Mars, including

government contracts, and other and multiplanetary initiatives could be delayed, modified, or cancelled and could

materially and adversely affect our business, financial condition, and results of operations. Even if such goals are

achieved, they may not generate meaningful revenue or achieve profitability for an extended period of time.

Our AI segment is recently formed, is still being fully integrated and optimized, operates in a rapidly evolving

industry and is subject to significant execution, competitive and operational risks.

We acquired xAI in February 2026 as the foundational platform for our AI segment and as part of our ambitious

vertical integration strategy intended to combine artificial intelligence capabilities with our established Space and

Connectivity businesses. Prior to its acquisition by the Company, xAI itself was an early-stage company. As a result,

our AI segment remains in a relatively early stage of organizational and operational maturity and is subject to

integration, scaling and execution risks.

The successful integration of acquired businesses, technologies, strategic partners, and employees is inherently

complex, costly and time-consuming, and may result in operational inefficiencies, delays, disruptions, increased

costs, loss of knowledge and diversion of management attention. As is common in large acquisitions, we have had to

take significant steps to integrate xAI’s operations into our broader corporate structure as part of our AI segment,

including putting in place the management team and organizational structure needed to execute at the scale and pace

our strategy demands, as well as controls and procedures appropriate for a larger organization like ours. Many of

these steps are not yet complete.

We have undertaken, and continue to undertake, changes in personnel, strategic partnerships, infrastructure-sharing

arrangements, organizational restructurings, acquisitions and other integration initiatives intended to accelerate

development of our AI capabilities, compute infrastructure and commercial offerings. Management believes these

initiatives may create long-term strategic advantages through the combination of engineering talent, compute

infrastructure, proprietary data, software capabilities and integrated operational platforms across the Company’s

businesses, among others. However, the successful integration of acquired businesses, management teams,

employees, strategic partners, technologies and evolving product architectures is inherently complex, costly and

time-consuming and may result in operational inefficiencies, delays, disruptions or the failure to realize anticipated

synergies or commercial benefits.

We have also pursued evolving commercial and technical strategies, including coding and software development

(such as through our partnership with Cursor) and monetization of unused compute capacity (such as through our

cloud compute services agreements with Anthropic), while simultaneously continuing to invest heavily in expanding

datacenter and compute capacity for our own internal AI initiatives and products. These efforts may require

substantial capital expenditures and management attention and may create operational complexity relating to

infrastructure allocation, prioritization of internal versus external compute usage, integration of third-party

technologies and partnerships, cybersecurity, data governance and commercialization strategy. We may elect to

allocate capital and resources to long-term initiatives even if alternative uses with more short-term upside are

available. There can be no assurance that these initiatives will achieve their intended operational or financial

objectives.

54

The artificial intelligence industry is highly dynamic and rapidly evolving. We face significant uncertainty relating

to technological developments, changing customer preferences, evolving regulatory and legal frameworks,

increasing public scrutiny, and intense competition for engineering talent, compute capacity, infrastructure,

customers and capital. In addition, the consumer AI market is characterized by rapid model iteration, frequent new

entrants and intense competition for user attention; as a result, download and other usage metrics for any individual

AI application, including Grok, can fluctuate significantly (including periods of decreased Grok app downloads and

user activity) in response to competitor model releases, product update cycles, and broader shifts in user behavior.

As a result of these market dynamics, we may need to modify our AI strategy, organizational structure,

infrastructure deployment and capital allocation decisions in response to technological change, competitive

pressures, regulatory developments or commercial adoption trends. Initiatives that management believes are

strategically beneficial over the long term may nevertheless experience near-term operational disruptions, integration

inefficiencies, product delays, technical setbacks, leadership turnover, employee attrition, infrastructure constraints,

increased costs or uneven customer adoption during periods of transition or rapid scaling.

Management believes that our recent organizational restructuring efforts, infrastructure investments and strategic

collaborations position the AI segment favorably for long-term growth and are consistent with the maturation

process of rapidly scaling AI platforms and optimization of acquired companies. However, there can be no assurance

that we will successfully integrate acquired businesses and technologies, retain key personnel, execute our AI

strategy within anticipated timeframes, achieve meaningful commercial adoption, generate anticipated revenues or

returns on investment, or compete effectively in a rapidly evolving and increasingly competitive and consolidated AI

market. If we are unable to successfully execute our AI strategy, our business, financial condition, results of

operations and prospects could be materially adversely affected.

Our AI segment is capital intensive, has incurred significant operating losses, and operates in a nascent and

rapidly evolving market in which the potential of AI remains uncertain.

AI is a nascent and rapidly evolving technology, and although we believe AI holds significant promise for

consumers and enterprises, its long-term impact will depend on the degree to which AI products and services prove

to be broadly useful in real-world applications. There can be no assurance that demand for AI solutions will develop

or be sustained at the levels we anticipate, or at all. While industry interest in AI has grown substantially, the

commercial value proposition of frontier AI models remains largely unproven, and long-term market acceptance of

our AI products and services is uncertain. Developing, training, and providing inference for frontier AI models

requires substantial and growing capital expenditures, including investments in specialized computing hardware,

data center infrastructure, energy procurement, and technical personnel, and we expect these costs to continue to

increase for the foreseeable future. In addition, we plan to allocate substantial capital to build our AI compute

infrastructure, and we expect a multi-year investment horizon before these deployments translate into sustained

positive AI Segment Adjusted EBITDA. Our AI segment has incurred significant operating losses since inception,

and we may not achieve profitability in this segment, or, if achieved, sustain it, and there can be no assurance that

the returns on our AI investments will be adequate to justify the capital deployed. Furthermore, the continued

improvement of AI model capabilities has historically depended in part on scaling laws, the empirical observation

that model performance improves with increased compute, data, and model size, but there is uncertainty as to how

long these scaling relationships will continue to hold. As a result of these factors, our AI segment may not achieve

the growth or returns we expect.

We have a history of net losses and may not achieve profitability in the future.

We incurred net losses of $(4,937) million and $(4,628) million for the years ended December 31, 2025 and 2023,

respectively, and a net loss of $(4,276) million for the three months ended March 31, 2026. We may not achieve or,

if achieved, sustain profitability in the future. As of March 31, 2026, we had an accumulated deficit of $41,311

million. While we have experienced significant growth in revenue over the last three years, we cannot predict

whether we will maintain this level of growth or when we will achieve profitability again. We also expect our capital

expenditures and operating expenses to increase in the future, including our general and administrative expenses as a

result of increased costs associated with operating as a public company and as we continue to invest for our future

growth, including substantial capital expenditures to design, develop, expand, and maintain our technologies and

infrastructure to support our operations. Our revenue could decline for a number of reasons, including if we are

55

unable to execute on our growth strategy and as a result of the other risks described in this prospectus. Furthermore,

if we fail to maintain or increase our revenue to offset increases in our operating expenses or manage our costs as we

invest in our business, including if we do not maintain or improve our operating efficiencies, we may not achieve or

sustain profitability. Any failure by us to achieve or sustain profitability on a consistent basis could have a material

adverse effect on our business, financial condition and results of operations and cause the market price of our Class

A common stock to decline.

The timing of our revenue and cost recognition may fluctuate due to factors outside of our control, which could

cause our periodic results of operations to fluctuate and make our results difficult to predict.

In our financial results, we recognize revenue and costs for a majority of customer payloads at the launch or

deployment of the customer’s payload to its intended orbit. While we plan launches and schedule payloads in

advance, the timing of these launches or deployments may vary and can be delayed or otherwise affected by a

number of factors outside of our control, including the customer’s delay in delivering their payload for integration

onto the launch vehicle, adverse weather, and other operational considerations. As a result, the timing of revenue

recognition may shift between reporting periods. For example, if the launch of a customer’s payload was expected to

occur near the end of a reporting period but instead occurs shortly thereafter (e.g., on April 1 instead of March 30),

the associated revenue would be recognized in the subsequent quarter. In addition, if a significant number of

launches or deployments occur within a short period of time, the concentration of those events may result in greater

variability in the timing of revenue recognition between reporting periods. These factors may cause our quarterly or

annual results of operations to fluctuate and may make our results difficult to predict.

Failure to comply with requirements to design, implement, and maintain effective internal controls could have a

material adverse effect on our business and stock price.

As a privately held company, we were not required to evaluate our internal control over financial reporting in a

manner that meets the standards of publicly traded companies required by Section 404(a) of the Sarbanes-Oxley Act

(“Section 404”).

As a public company, we will have significant requirements for enhanced financial reporting and internal controls.

The process of designing and implementing effective internal controls is a continuous effort that requires us to

anticipate and react to changes in our business and the economic and regulatory environments and to expend

significant resources to maintain a system of internal controls that is adequate to satisfy our reporting obligations as

a public company. If we are unable to establish or maintain appropriate internal financial reporting controls and

procedures, it could cause us to fail to meet our reporting obligations on a timely basis, result in material

misstatements in our consolidated financial statements, and harm our results of operations. In addition, we will be

required, pursuant to Section 404, to furnish a report by management on, among other things, the effectiveness of

our internal control over financial reporting in the second annual report following the completion of this offering.

This assessment will need to include disclosure of any material weaknesses identified by our management in our

internal control over financial reporting. The rules governing the standards that must be met for our management to

assess our internal control over financial reporting are complex and require significant documentation, testing, and

possible remediation. Testing and maintaining internal controls may divert our management’s attention from other

matters that are important to our business. Additionally, our independent registered public accounting firm will be

required to attest to the effectiveness of our internal control over financial reporting on an annual basis, beginning

with our second annual report.

We are currently in the process of updating our control processes and automating certain of our procedures and

systems in anticipation of becoming a public company, but our internal controls over financial reporting currently do

not meet all of the standards contemplated by Section 404 that we will eventually be required to meet. Because we

currently do not have comprehensive documentation of our internal controls and have not yet tested our internal

controls in accordance with Section 404, we cannot conclude in accordance with Section 404 that we do not have a

material weakness in our internal controls or a combination of significant deficiencies that could result in the

conclusion that we have a material weakness in our internal controls. In connection with updating our control

processes and the implementation of the necessary procedures and practices related to internal control over financial

reporting, we have identified deficiencies and may identify deficiencies in the future that we may not be able to

56

remediate in time to meet the deadline imposed by the Sarbanes-Oxley Act for compliance with the requirements of

Section 404. In addition, we may encounter problems or delays in completing the remediation of any deficiencies

identified by our independent registered public accounting firm in connection with the issuance of their attestation

report. Our testing, or the subsequent testing (if required) by our independent registered public accounting firm, may

reveal deficiencies in our internal controls over financial reporting that are deemed to be material weaknesses. Any

material weaknesses could result in a material misstatement of our annual or quarterly consolidated financial

statements or disclosures that may not be prevented or detected.

Our insurance coverage strategy may not be adequate to protect us from all business risks.

We may be subject, in the ordinary course of business, to losses resulting from accidents, acts of God and other

claims against us, for which we may have no insurance coverage. As a general matter, we do not maintain as much

insurance coverage as many other companies do, and in some cases, we do not maintain any at all, including with

respect to our in-orbit satellites, which we currently do not insure and do not expect to insure in the future.

Additionally, the policies that we do have may include significant deductibles or self-insured retentions, policy

limitations and exclusions, and we cannot be certain that our insurance coverage will be sufficient to cover all future

losses or claims against us. A loss that is uninsured or which exceeds policy limits may require us to pay substantial

amounts, which may harm our financial condition and operating results.

Risks Related to Our Corporate Structure, Ownership of our Class A Common Stock and This Offering

Conflicts of interest could arise in the future between us, on the one hand, and Mr. Musk and entities owned by

or affiliated with him, on the other hand, concerning among other things, business transactions, potential

competitive activities or other business opportunities.

Conflicts of interest could arise in the future between us, on the one hand, and Mr. Musk and entities owned by or

affiliated with him, on the other hand, concerning among other things, business transactions, potential competitive

business activities or other opportunities. In the normal course of business, we have engaged in a variety of

transactions with some of these companies. Please refer to “Certain Relationships and Related Person Transactions.”

In addition, we have previously engaged, are currently engaged, and expect to continue to engage in the future in a

number of strategic collaborations with Tesla, including with respect to Macrohard and Terafab. Certain of these

projects, including Macrohard and Terafab, are in the very early stages, as a result of which we and Tesla have not

finalized a variety of details relating to our collaboration, including, but not limited to, financial terms, intellectual

property rights, and the ultimate term of our collaboration. Furthermore, Mr. Musk and other businesses owned by

or affiliated with him may now, or in the future, directly or indirectly, compete with us for investment or business

opportunities.

Mr. Musk or his affiliates may become aware, from time to time, of certain business opportunities (such as

acquisition opportunities or technological developments) and may direct such opportunities to other businesses in

which they have invested, in which case we may not become aware of or otherwise have the ability to pursue such

opportunity. In addition, Mr. Musk and his affiliates may dispose of their interests in other companies or other assets

in the future, without any obligation to offer us the opportunity to purchase any of those interests or assets.

Under our charter, Mr. Musk and his affiliates are not restricted from owning assets or engaging in businesses that

compete directly or indirectly with us and will not have any duty to refrain from engaging, directly or indirectly, in

the same or similar business activities or lines of business as us, including those business activities or lines of

business deemed to be competing with us, or doing business with any of our customers or vendors. Moreover, we

have in the past entered into, and may in the future enter into, transactions with entities affiliated with Mr. Musk. We

may enter into such transactions in lieu of pursuing other opportunities that some other shareholders may prefer or

that may prove to be more accretive than the opportunities we elect to pursue. In any of these matters, the interests

of Mr. Musk and entities owned by or affiliated with him may differ or conflict with the interests of our other

shareholders. Any actual or perceived conflicts of interest with respect to the foregoing could have an adverse

impact on the trading price of our Class A common stock.

57

Certain of our directors and key employees may have conflicts of interest because they are also employees or

directors of affiliates of Mr. Musk or other large shareholders. The resolution of these conflicts of interest may

not be in our or your best interests.

Certain of our directors and key employees may have conflicts of interest because they are also employees or

directors of affiliates of Mr. Musk or other large shareholders. Such directors may have interests in, serve on the

boards of, or have financial or other relationships with other companies, ventures, or initiatives that are related to or

competitive with our business, including but not limited to other space or AI companies, technology ventures,

satellite communications businesses, and government or commercial space contracts. Please refer to “Management.”

These relationships and interests could create actual or perceived conflicts of interest, particularly with respect to the

allocation of time, resources, business opportunities, or strategic decisions. In addition, our charter provides that, to

the fullest extent permitted by applicable law, we renounce certain corporate opportunities that may be presented to

Mr. Musk and certain of our directors and their respective affiliates, and such persons may have no duty to present

such opportunities to us. Please refer to “Description of Capital Stock—Corporate Opportunities.” Any actual or

perceived conflicts of interest could harm our reputation, lead to disputes, divert management attention, or result in

decisions that are not in the best interests of us or our shareholders, which could materially and adversely affect our

business, financial condition, results of operations, and future prospects.

We are highly dependent on the continued services of Mr. Musk, our Chief Executive Officer and Chief

Technical Officer, and other key personnel, and the loss or reduced involvement of one or more of these

individuals could adversely affect our ability to execute our business strategy.

We are highly dependent on the continued service and performance of Mr. Musk, whose leadership, vision, and

expertise are critical to the development of our technologies and the execution of our business strategy. Mr. Musk

has been, and continues to be, a driving force behind our growth, innovation, and operational success. The loss of

Mr. Musk, whether due to death, disability, or otherwise, or his inability or unwillingness to continue in his current

roles, could significantly disrupt our management structure, adversely affect our ability to execute our strategic

plans, and negatively impact our reputation and relationships with customers, partners, and other stakeholders. Our

intense, mission-driven, engineering-first culture has been a key driver of our growth and execution, and any erosion

of this culture, including as a result of the loss or reduced involvement of Mr. Musk, could have a material adverse

effect on our business, financial condition, results of operations, and future prospects. We do not maintain key-

person life insurance on Mr. Musk. Further, although Mr. Musk devotes significant time to our businesses and is

highly active in our management, he does not devote his full time and attention to our businesses and devotes time

and attention to other significant roles (and may in the future serve in additional roles). For instance, Mr. Musk

currently serves as Technoking and Chief Executive Officer of Tesla and is involved in other emerging technology

ventures, including Neuralink and The Boring Company. Mr. Musk has also previously served as Senior Advisor to

the President of the United States. Any such loss or reduced involvement in our business could result in a material

adverse effect on our business, financial condition, results of operations, and future prospects. The process of

identifying and recruiting a successor with the combination of skills and experience possessed by Mr. Musk, as well

as the ability to maintain the confidence of the market, could be lengthy and uncertain, and there can be no assurance

that we would be able to attract or retain a suitable replacement in a timely manner or at all.

We, Mr. Musk, and other companies Mr. Musk is affiliated with frequently receive an immense amount of media

attention. The actions and statements of Mr. Musk and his affiliated ventures, whether or not directly relating to us,

may draw significant public attention and scrutiny to us and could potentially have a positive or negative impact on

our business, relationships with customers and regulators, or stock price.

In addition to Mr. Musk, we have key personnel who are invaluable to our businesses. We rely upon their

knowledge, expertise, and leadership to develop, manufacture, launch, sell, and support our products and services.

None of our key employees are bound by an employment agreement for any specific term and we may not be able to

successfully attract and retain the senior leadership necessary to continue to grow our business. Our compensation

arrangements, such as our equity award programs, may not always be successful in attracting new employees and

retaining and motivating existing key personnel. Our success depends upon our ability to attract and retain key

personnel and any failure to do so could have a material adverse effect on our business, financial condition, results

of operations, and future prospects.

58

A significant reduction by Mr. Musk or other existing shareholders of their ownership interest in us could

adversely affect us.

We believe that Mr. Musk’s substantial ownership interest in us provides him with an economic incentive to assist

us to be successful. Upon the expiration or earlier waiver of the lock-up restrictions on transfers or sales of our

securities following the completion of this offering, Mr. Musk will not be subject to any obligation to maintain his

ownership interest in us and may elect at any time thereafter to sell all or a substantial portion of or otherwise reduce

his ownership interest in us. If Mr. Musk sells all or a substantial portion of his ownership interest in us, he may

have less incentive to assist in our success, which could adversely affect our future prospects. Additionally, future

resales of our Class A common stock by Mr. Musk or other existing shareholders, or the perception that such sales

may occur, could cause the market price of our Class A common stock to decline significantly, regardless of our

actual business performance. In particular, subject to the expiration or waiver of any applicable lock-up period,

parties to the Investors’ Rights Agreement described in “Certain Relationships and Related Person Transactions—

Investors’ Rights Agreement” will have the right, subject to certain exceptions and conditions, to require us to

register approximately shares of Class A common stock under the Securities Act, and they will have the

right to participate in certain future registrations of securities by us. Registration of any of such shares would result

in such shares becoming freely tradable without compliance with Rule 144 limitations upon effectiveness of the

registration statement. In addition, approximately shares of Class A common stock will generally be

available for resale under Rule 144 starting 90 days after this offering, subject to lock-up restrictions described

elsewhere in this prospectus. See “Certain Relationships and Related Person Transactions—Investors’ Rights

Agreement” and “Shares Eligible for Future Sale—Registration Rights.”

Following the consummation of this offering, we will be a “controlled company” within the meaning of the

Nasdaq and Nasdaq Texas listing rules and, as a result, will qualify for and rely on exemptions from certain

corporate governance requirements.

Because Mr. Musk will beneficially own                    shares of Class A common stock and                    shares of

Class B common stock, which represents greater than 50% of the voting power of our common stock with respect to

director elections and moreover, holders of our Class B common stock, voting separately as a class, will be entitled

to elect 51% of the total number of authorized directors constituting our board (rounded up to the nearest whole

number), following the completion of this offering, we will be a controlled company under the listing rules of

Nasdaq and Nasdaq Texas.

Under the listing rules of Nasdaq and Nasdaq Texas, a company of which more than 50% of the voting power with

respect to director elections is held by another person or group of persons acting together is a “controlled company”

and may elect not to comply with certain Nasdaq and Nasdaq Texas corporate governance requirements, including

the requirements that:

•a majority of such company’s board of directors consist of independent directors as defined under the listing

rules of Nasdaq and Nasdaq Texas;

•director nominees be selected or recommended for board of directors’ selection by a nominating committee

composed entirely of independent directors, with a written charter addressing the nominations process as

required under the listing rules of Nasdaq and Nasdaq Texas;

•the compensation committee be composed entirely of independent directors with a written charter addressing

the committee’s purpose and responsibilities; and

•annual performance evaluations of the compensation and nominating committees be conducted.

Following the completion of this offering, we intend to utilize certain of these exemptions. As a result, we do not

expect to have a compensation and nominating committee that is composed entirely of independent directors or that

has a committee charter that addresses all Nasdaq and Nasdaq Texas requirements applicable to companies that are

not controlled companies. Additionally, we may elect to take advantage of certain other exemptions in the future for

as long as we remain a “controlled company.” Accordingly, our Class A shareholders will not have the same

protections afforded to shareholders of companies that are subject to all of the corporate governance requirements of

59

Nasdaq and Nasdaq Texas. In the event that we cease to be a “controlled company” and our shares continue to be

listed on Nasdaq and Nasdaq Texas, we will be required to comply with all of the applicable governance

requirements within the applicable transition periods. Please refer to “Management.”

Our ability to provide returns to shareholders will depend on appreciation in our share price, as we do not plan to

pay dividends for the foreseeable future.

The ability of investors to realize a return on their investment will depend largely on the appreciation of the price of

our Class A common stock, as we do not anticipate paying dividends in the foreseeable future. We have never

declared or paid any cash dividends on our common stock, and we currently intend to retain all available funds and

any future earnings to support the growth and operation of our business, including investment in new technologies

and commercial opportunities. As a result, investors seeking cash returns from their investment will not receive any

dividend income, and the only way to realize a return may be through an increase in the market price of our Class A

common stock, which may not occur. The trading price of our Class A common stock may be volatile and subject to

wide fluctuations in response to various factors, including our financial condition and operating results, changes in

our business or future prospects, technological innovations, announcements by us or our competitors, changes in the

regulatory environment, harm to our brand and reputation, broader market or economic conditions, and the fact that

a number of shares of our Class A common stock are expected to be allocated to retail investors in this offering.

Additionally, high retail investor interest in our Class A common stock may occur following this offering, which

may lead to increased volatility of the trading price. Some of these factors are outside of our control, and the trading

price of our Class A common stock may not reflect our actual operating performance. Accordingly, investors may

not be able to realize a gain on their investment and could lose all or part of their investment in our Class A common

stock.

Upon completion of this offering, Mr. Musk will serve as our Chief Executive Officer, Chief Technical Officer,

and Chairman of our board and control the election of our directors, and our dual class structure concentrates

voting control with Mr. Musk and other holders of our Class B common stock. This will limit or preclude your

ability to influence corporate matters and the election of our directors.

Our Class B common stock will have ten votes per share; our Class A common stock will have one vote per share;

and, except as summarized here, our Class A common stock will vote together with our Class B common stock on

any matter submitted to the shareholders for a vote. Under our charter, holders of our Class B common stock, voting

separately as a class, will be entitled to elect 51% of the total number of authorized directors constituting our board

(rounded up to the nearest whole number) and will have the ability to remove those directors for as long as there is at

least one share of Class B common stock outstanding. As a result, holders of our Class B common stock will have

control over the composition of our board and significant influence over the outcome of matters requiring

shareholder approval. Please refer to “Description of Capital Stock” for certain other actions that will require

approval of a majority of the voting power of the outstanding shares of Class B common stock voting separately as a

class. This concentration of voting power will limit or preclude the ability of holders of our Class A common stock,

including purchasers of Class A common stock in this offering, to influence corporate matters and the election of our

directors.

Upon completion of this offering, Mr. Musk will beneficially own a majority of the outstanding shares of our

Class B common stock and a majority of the voting power of the common stock (the Class A common stock and the

Class B common stock voting together) and therefore will be able to elect all the members of our board. Mr. Musk,

who will serve as our Chief Executive Officer and Chairman of our board under our charter and can only be

removed from our board or these positions by the vote of Class B holders, as set forth in our charter, will exert

significant influence over our business and affairs.

Class B common stock will continue to have ten votes per share, except that, subject to exceptions for certain inter-

family transfers and transfers to certain entities that qualify as “permitted transferees” (as described elsewhere in this

prospectus), transfers by holders of our Class B common stock will generally result in those shares converting to

Class A common stock. The conversion of Class B common stock to Class A common stock will have the effect,

over time, of increasing the relative voting power of those holders of Class B common stock who retain their shares.

60

If Mr. Musk retains a significant portion of his holdings of Class B common stock for an extended period of time, he

could continue to control the election and removal of a majority of our board.

However, other persons will also hold shares of Class B common stock. If Mr. Musk were to sell, transfer or

otherwise dispose of a sufficient number of his shares of Class B common stock such that he no longer holds a

majority of the outstanding shares of Class B common stock, another holder or group of holders of Class B common

stock could obtain the ability to elect and remove a majority of our board and thereby effectively control the

Company. Any such change in control could result in changes to our strategic direction, management, business plans

or policies that may not be aligned with the interests of holders of our Class A common stock.

In addition, our charter will provide that other than for specified class votes by the Class B common stock or any

rights granted to other classes in the future, classes of stock will not be entitled to any separate class votes provided

for under the Texas Business Organizations Code (the “TBOC”), including among others (i) the increase or decrease

of the aggregate number of authorized shares of a class outstanding, (ii) the exchange, reclassification, or

cancellation of all or part of the shares of a class, (iii) a change of shares of a class, with or without par value, into

the same or a different number of shares of the same or another class, with or without par value, (iv) the creation of a

new class of shares with rights and preferences equal, prior, or superior to the shares of the class and (v) cancellation

or other effectuation of the dividends on the shares of the class or series that have accrued but have not been

declared.

The TBOC and our charter include provisions that may limit shareholders’ ability to bring a cause of action

against our directors or officers for certain acts or omissions in their capacity as directors or officers of the

Company, including minimum share ownership for derivative proceedings and the presumption of the business

judgment rule.

The TBOC and our governing documents include certain provisions that may limit our shareholders’ ability to bring

certain derivative claims against our officers and directors. For example, the TBOC provides that, if a corporation

has a class of stock listed on a national securities exchange, the governing documents may provide that the minimum

ownership threshold for a shareholder or group of shareholders to institute or maintain such derivative proceeding is

3% of shares outstanding. A similar ownership threshold provision based on this 2025 TBOC provision has already

been challenged in court proceedings involving another Texas corporation and, although the federal district court

found the provision enforceable in that case, its enforceability or governing documents containing its provisions

could be subject to further challenges or interpretation. The TBOC also permits corporations to request a court, at

the start of a transaction (including a related party transaction) or inquiry into a derivative claim, to determine the

independence and disinterestedness of directors serving on a special committee reviewing the transaction or

directors or other individuals on panels reviewing derivative claims. Subsequent challenges to independence or

disinterestedness would require new facts. Our bylaws will provide that these TBOC provisions will apply to us.

In addition, Section 21.419 of the TBOC sets forth certain presumptions concerning compliance by directors and

officers with respect to their duties to a corporation, including the duty of care and duty of loyalty. Specifically, in

taking or declining to take any action on any matters of a corporation’s business, Section 21.419, which applies to

us, provides that a director or officer is presumed to have acted (i) in good faith, (ii) on an informed basis, (iii) in

furtherance of the interests of the corporation and (iv) in obedience to the law and the corporation’s governing

documents. These provisions are described as codifying the “business judgment rule.” In order to succeed in a cause

of action against a director or officer, the Company or a shareholder pursuing such an action must rebut one or more

of the foregoing presumptions and prove with particularity the director or officer’s act or omission constituted a

breach of duty as a director or officer and that such breach involved fraud, intentional misconduct, an ultra vires act

or a knowing violation of law.

Our bylaws will impose minimum stock ownership and solicitation requirements on shareholders seeking to

submit proposals for shareholder approval, which could limit the ability of our shareholders to bring matters

before a meeting of shareholders.

Upon the completion of this offering, we will qualify as a “nationally listed corporation” under Section 21.373 of the

TBOC, and our bylaws will provide that the shareholder proposal requirements permitted by that section will apply

61

immediately upon qualifying as a “nationally listed corporation.” As a result, except with respect to director

nominations and procedural resolutions ancillary to the conduct of a shareholders’ meeting, a shareholder or group

of shareholders seeking to submit a proposal for approval at a meeting of shareholders will be required to satisfy

specified ownership, holding-period and solicitation requirements. Under these provisions, the proposing

shareholder or shareholder group must hold an amount of voting shares (determined as of the date of submission of

the proposal) equal to at least 3% of our voting shares, must have held that amount continuously for at least six

months before the date of the meeting and throughout the entire duration of the meeting, and must solicit holders of

shares representing at least 67% of the voting power of shares entitled to vote on the proposal at the shareholder

meeting. For the purpose of this paragraph, “voting shares” means shares that entitle the holder of the shares to vote

on the proposal. These requirements are more restrictive than the requirements that would otherwise apply absent

such a bylaw provision and may make it more difficult, or in some cases impracticable, for shareholders to submit

proposals for consideration at a shareholders’ meeting. As a result, our shareholders may have fewer opportunities to

present proposals for shareholder approval, even on matters they believe are important, which could limit

shareholder influence over corporate governance and other matters. Section 21.373 of the TBOC was enacted in

2025 and, while its enforceability has not yet been challenged in court and we do not have any material concerns

related to enforceability of Section 21.373 or the related bylaws provision, like many new laws, we expect the

enforceability of TBOC Section 21.373 will eventually be challenged.

Our bylaws place restrictions on the forum, venue and procedures for legal actions or proceedings initiated by

our shareholders, including certain requirements for mandatory arbitration. These provisions could limit our

shareholders’ ability to pursue certain claims and/or increase the cost of doing so and could also affect the

procedures, rights, and remedies available to our shareholders in such legal actions or proceedings.

Our bylaws will contain a section (the “Forum Section Bylaw”) that will provide that, unless the Company consents

in writing to the selection of an alternative forum, the sole and exclusive forum for the filing, adjudication, and trial

of all disputes between (i) one or more shareholders and (ii) the Company or its directors, officers, or controlling

persons, or any underwriter of securities issued by the Company (or controlling person thereof) relating to any of the

following: (1) any derivative proceeding, meaning a civil dispute brought in the right of the Company; (2) any action

based on the governance, governing documents, or internal affairs of the Company; (3) any action based on state or

federal securities or trade regulation laws; (4) any action based on the alleged act(s) or omission(s) by a person in its

capacity as a shareholder, controlling person, director, officer or other managerial official of the Company; (5) any

action based on the alleged breach(es) by one or more shareholders, controlling persons, directors, officers, or other

managerial officials of a duty owed, in his or her capacity as such, to the Company or to any shareholder thereof; (6)

an action seeking to hold a shareholder, controlling person, director, officer, or other managerial official of the

Company liable for an obligation of the Company, other than on account of a written contract signed by the person

to be held liable in a capacity other than as a shareholder or managerial official; and (7) any action arising out of the

TBOC, will be the Texas Business Court, Eleventh Division (the “Business Court”) (for purposes of this summary,

each, an “Internal Dispute”).

The selection of the Business Court as the exclusive forum for Internal Disputes may limit a shareholder’s ability to

bring a claim in a judicial forum that it finds favorable for disputes with us or our directors, officers, other

managerial officials, or other employees, which may discourage lawsuits against us and our directors, officers, other

managerial officials, and other employees. Except to the extent that the Company consents in writing, or a court of

competent jurisdiction determines in a final and unappealable judgment, that an Internal Dispute is not subject to the

sole and exclusive venue and forum or jurisdiction of the Business Court or arbitration (as described further below),

a shareholder will not be permitted to litigate an Internal Dispute in federal court or in any state court other than the

Business Court, and will not be able to avail itself of any potential advantages or procedural protections of such

other forums. Any person or entity purchasing or otherwise acquiring any interest in our shares of capital stock will

be deemed to have notice of and have consented to these provisions. For more information, please refer to

“Description of Capital Stock—Anti-Takeover Effects of Provisions of Our Charter, our Bylaws and Texas Law.”

SpaceX maintains that the Forum Selection Bylaw, including without limitation the selection of the Business Court

as the sole and exclusive forum for all actions brought under federal securities laws, accords with the law and is

enforceable. However, the law governing the selection of a forum other than a federal court for certain actions

brought under the federal securities laws is unsettled, and there is some risk that, if an Internal Dispute were filed

62

under the Exchange Act (or the rules and regulations thereunder) in a court other than the Business Court, that court

could deny a motion to transfer the action to the Business Court pursuant to the Forum Selection Bylaw.

Accordingly, the bylaws provide that to the extent that a court of competent jurisdiction were to determine in a final

and unappealable judgment that an Internal Dispute is not subject to the sole and exclusive venue and forum or

jurisdiction of the Business Court (such Internal Dispute, an “Other Dispute”), such Other Dispute would be

exclusively and finally settled by arbitration, pursuant to the Texas Arbitration Act, under the Expedited Procedure

Provisions of the Rules of the International Chamber of Commerce, pursuant to Article 30 thereof. To be clear,

absent Company consent, a shareholder would not be able to file an arbitration demand pursuant to the Dispute

Resolution Clause without first obtaining a final and unappealable judgment that the shareholder’s Internal Dispute

is not subject to the sole and exclusive venue and forum or jurisdiction of the Business Court. The governing law of

such Other Dispute would be the federal law of the United States or the law of the State of Texas, as applicable to

the issues raised in the Other Dispute, including without limitation the pleading and discovery limitations of the

Private Securities Litigation Reform Act.

Given the unsettled nature of the law in this area, there is also some risk that a court that has denied a motion to

transfer an Internal Dispute to the Business Court pursuant to the Forum Selection Bylaw would also deny a motion

to compel arbitration of such Other Dispute pursuant to the Forum Selection Bylaw. Accordingly, the Forum

Selection Bylaw further provides that to the extent that a court of competent jurisdiction determines in a final and

unappealable judgment that such Other Dispute cannot be compelled to arbitration pursuant to the Forum Selection

Bylaw, the sole and exclusive forum for the adjudication and trial of such Other Dispute will be the United States

District Court for the Southern District of Texas, Houston Division (the “Federal Court”).

Finally, the Forum Selection Bylaw provides that to the extent that a court of competent jurisdiction determines in a

final and unappealable judgment that the Federal Court lacks jurisdiction over any such Other Dispute, the sole and

exclusive forum and venue for such Other Dispute will be the state district courts of Harris County, Texas.

Regardless of the forum, venue, or procedures selected for an Internal Dispute or Other Dispute, our bylaws shall

require that any Internal Dispute or Other Disputes be brought only as an individual action or derivative proceeding,

and, to the fullest extent permitted by law, shall prohibit shareholders from bringing such an Internal Dispute or

Other Dispute as a class action, mass action, or other form of collective action or from being consolidated or joined,

in whole or in part, consistent with the Arbitration Rules. However, the Company, at its sole option, may elect to

seek consolidation or joinder of matters as consistent with the Arbitration Rules.

In addition, our bylaws will provide that any person or entity purchasing or otherwise acquiring or holding any

interest in shares of stock of the Company shall be deemed to have irrevocably and unconditionally waived any right

it may have to a trial by jury in any Internal Dispute. This will prevent a shareholder from requesting that a jury

decide disputed issues of fact and may discourage lawsuits against us and our directors, officers, other managerial

officials, and other employees.

These dispute resolution rules that our bylaws will establish for Internal Disputes, as well as the Arbitration Rules to

the extent they will apply, are different from the procedural rules that would normally apply to the litigation of

Internal Disputes in state or federal court. They may prevent a shareholder from availing itself of procedural

protections that would be available under litigation in state or federal court and may render available or affect

adversely the rights and remedies available to shareholders in such proceedings. Particularly in the case of

arbitration, including its prohibition on class or collective actions, these dispute resolution rules may also result in

greater costs being imposed on shareholders to litigate Internal Disputes, and in some cases involving lower amounts

in controversy, the additional costs that may be imposed on shareholders to litigate Internal Disputes could exceed

the potential recovery from such litigation.

It is possible that one or more provisions of our bylaws, including those regarding the exclusive forum for Internal

Disputes, mandatory arbitration for Other Disputes, or waiver of the right to proceed on a class, mass, or collective

basis, may be found by a court to be inapplicable or unenforceable. In addition, the mandatory arbitration provision

in our bylaws could be subject to litigation or regulatory scrutiny, which could result in the provision being enjoined

or in additional costs or uncertainty. In such case, we may incur additional costs or delays associated with resolving

63

such actions, including in other jurisdictions, which could adversely affect our business, financial condition, or

results of operations.

64

CAUTIONARY STATEMENT REGARDING FORWARD-LOOKING STATEMENTS

This prospectus contains forward-looking statements. Forward-looking statements include those that express a

belief, expectation, or intention, as well as those that are not statements of historical fact. Forward-looking

statements contained in this prospectus include information regarding our future operating results and financial

position, our business strategy and plans and our objectives for future operations. Forward-looking statements

contained in this prospectus also include, but are not limited to, statements about:

•the development and deployment of Starship in accordance with our anticipated schedule (including

commencement of payload delivery to orbit in 2026) and launch cadence and our ability to achieve expected

performance, reusability, and cost efficiencies;

•the size and growth of our various existing and future markets, including the markets for commercial launch

services, satellite connectivity services, our AI platforms, AI compute infrastructure (terrestrial and orbital),

lunar-related activities and interplanetary activities, including the extent to which such markets develop,

particularly emerging or unproven markets that may not materialize as expected or on anticipated timelines;

•demand for our products and services, including our launch, connectivity, and AI offerings, and our ability to

grow our customer base and generate revenue;

•the deployment of our next-generation Starlink satellites, satellite-to-mobile connectivity, and orbital AI

compute infrastructure (including potential deployment of our orbital AI compute satellites as early as 2028),

including our ability to successfully develop, scale, and commercialize such technologies, which are subject to

significant technical complexity, capital requirements, new innovations and regulatory approvals;

•our target launch cadence and expansion of our manufacturing and operational capacity necessary to support our

strategies, including our ability to scale production, supply chain, infrastructure, and workforce efficiently;

•our ability to execute our growth strategy and scale our operations efficiently, including managing costs,

timelines, and operational complexity;

•our ability to solve novel issues and navigate and monetize technologies and environments that have never been

accessed or economized before;

•our ability to design, develop and successfully commercialize new and innovative technologies, products, and

services, including our AI platforms and Terafab, and our ability to achieve and maintain a low cost per token,

in each case in rapidly evolving and competitive markets;

•our ability to scale and monetize our AI products and services, including the development, performance, and

adoption of our frontier models and related applications, and to realize benefits from related acquisitions and

initiatives, such as our arrangement with Cursor;

•the amount, nature and timing of our capital expenditures and the impact of such capital expenditures on our

growth and performance, including our ability to fund such expenditures, manage costs, strategically reduce

costs and achieve expected returns on investment;

•our ability to obtain sufficient power, GPUs, and other critical components and manage our supply chain to

support our operations and growth;

•our ability to obtain and maintain required regulatory approvals, licenses and spectrum authorizations in the

United States and internationally, and the timing, scope, and conditions of such approvals;

•the competitive landscape in the industries in which we operate and our ability to compete effectively;

•the implementation, interpretation, and impact of current or future regulations including laws and regulations

relating to space operations, communications, AI, data privacy, and other areas;

•our ability to realize benefits and manage risks of being a public company; and

65

•general economic conditions.

These forward-looking statements may be accompanied by words such as “anticipate,” “believe,” “estimate,”

“expect,” “intend,” “may,” “outlook,” “plan,” “potential,” “predict,” “project,” “will,” “should,” “could,” “would,”

“likely,” “future,” “budget,” “goal,” “commit,” “pursue,” “target,” “seek,” “objective” or the negative of these

words, or similar expressions that are predictions of or indicate future events or trends that do not relate to historical

matters. We caution you that the foregoing list may not contain all of the forward-looking statements made in this

prospectus.

The forward-looking statements in this prospectus speak only as of the date of this prospectus, or such other date as

specified herein. We undertake no obligation to update these statements unless required by law, and we caution you

not to place undue reliance on them. Forward-looking statements are not assurances of future performance and

involve risks and uncertainties. We have based these forward-looking statements on our current expectations and

assumptions about future events. Forecasts, goals, milestones, and expectations that cover multi-year time horizons,

or unknown timelines, inherently involve increased risks with respect to predictability and actual results may differ

materially from current expectations. While our management considers these expectations and assumptions to be

reasonable, they are inherently subject to significant business, economic, competitive, regulatory, technological,

environmental, political, and other risks, contingencies and uncertainties, which are difficult to predict and many of

which are beyond our control. These risks, contingencies, and uncertainties and other important factors are described

in the “Risk Factors” and “Management’s Discussion and Analysis of Financial Condition and Results of

Operations” sections of this prospectus. Should one or more of such risks or uncertainties occur, or should

underlying assumptions prove incorrect, our actual results, performance, achievements or plans could differ

materially from those expressed or implied in any forward-looking statements. In addition, because we operate in

rapidly evolving and certain highly competitive markets, we may from time to time rapidly adjust, modify or change

our strategic priorities, capital allocation, product or service focus or operational initiatives in response to

technological developments, competitive dynamics, regulatory changes or other factors, which could cause actual

results to differ materially from those expressed or implied by the forward-looking statements contained herein. New

risks emerge from time to time, some risks are inherently unknown to us, and it is not possible for our management

to predict all such risks. Many of the risks and uncertainties that could materially adversely affect us or our prospects

are beyond our control or relate to portions of our business strategy that have a lengthy time horizon or involve

unprecedented ventures. This can make assessment of certain risks more difficult and you should factor these

uncertainties into your assessment of an investment in our Class A common stock. All forward-looking statements in

this prospectus are expressly qualified in their entirety by the cautionary statements in this section.

66

USE OF PROCEEDS

We expect to receive approximately $           of net proceeds from this offering (or $           if the underwriters

exercise their option to purchase additional shares of Class A common stock in full), based upon the assumed initial

public offering price of $           per share (which is the midpoint of the price range set forth on the cover page of this

prospectus) after deducting underwriting discounts and commissions and estimated offering expenses payable by us.

We intend to use the net proceeds from this offering to fund our growth strategy, including the expansion of our AI

compute infrastructure, enhancements to our launch infrastructure and launch vehicles, increases in the scale and

capacity of our satellite constellations, and any remaining amounts for general corporate purposes.

Assuming no exercise of the underwriters’ option to purchase additional shares, each $1.00 change in the assumed

initial public offering price of $           per share (which is the midpoint of the price range set forth on the cover page

of this prospectus) would cause the net proceeds from this offering, after deducting the underwriting discounts and

commissions and estimated offering expenses payable by us, to change by approximately $          million, assuming

no change to the number of shares of our Class A common stock offered by us, as set forth on the cover page of this

prospectus. Similarly, an increase (decrease) of one million shares of Class A common stock sold in this offering by

us would increase (decrease) our net proceeds by $          million, assuming the initial public offering price of

$           per share (which is the midpoint of the price range set forth on the cover page of this prospectus) remains

the same, and after deducting the underwriting discounts and commissions and estimated offering expenses payable

by us. If the net proceeds increase for any reason, we would use the additional net proceeds for the purposes set forth

above. If the net proceeds decrease for any reason, then we expect that we would use the lower amount of net

proceeds for the purposes set forth above.

The expected use of net proceeds from this offering represents our intentions based upon our present plans and

business conditions. We cannot predict with certainty all of the particular uses for the net proceeds from this offering

or the amounts that we will actually spend on each of the uses set forth above. Accordingly, our management will

have significant flexibility in applying the net proceeds from this offering. The timing and amount of our actual

expenditures will be based on many factors, including cash flows and the anticipated growth of our business.

67

DIVIDEND POLICY

We do not anticipate declaring or paying any cash dividends to holders of our common stock in the foreseeable

future. We currently intend to retain future earnings, if any, to finance the growth of our business. Our future

dividend policy is within the discretion of our board and will depend upon then-existing conditions, including our

results of operations, financial condition, capital requirements, investment opportunities, statutory restrictions on our

ability to pay dividends, restrictions in our existing and any future debt agreements and other factors our board may

deem relevant. Covenants under our Credit Agreements also restrict our ability to pay dividends, and we may enter

into credit agreements or other borrowing arrangements in the future that restrict our ability to declare or pay cash

dividends or make distributions in the future. Please refer to “Management’s Discussion and Analysis of Financial

Condition and Results of Operations—Liquidity and Capital Resources” for a description of the restrictions on our

ability to pay dividends.

Please refer to “Risk Factors—Risks Related to Our Corporate Structure, Ownership of our Class A Common Stock

and This Offering—Our ability to provide returns to shareholders will depend on appreciation in our share price, as

we do not plan to pay dividends for the foreseeable future.”

68

CAPITALIZATION

The following table sets forth our cash and cash equivalents and capitalization as of March 31, 2026:

•on an actual basis;

•on a pro forma basis, giving effect to (i) the Preferred Conversion as if such conversion had occurred on March

31, 2026, (ii) the Class C Reclassification as if such reclassification had occurred on March 31, 2026, and (iii)

the effectiveness of our charter, which will become effective upon the completion of this offering; and

•on a pro forma as adjusted basis, giving effect to (i) the pro forma adjustments set forth above, (ii) the sale of

shares of our Class A common stock in this offering at an assumed initial offering price of $           per share,

which is the midpoint of the range set forth on the cover page of this prospectus, and (iii) the application of the

net proceeds from this offering as described under “Use of Proceeds.”

The table below should be read in conjunction with, and is qualified in its entirety by reference to “Management’s

Discussion and Analysis of Financial Condition and Results of Operations,” “Description of Capital Stock” and our

consolidated financial statements and related notes included elsewhere in this prospectus.

	As of March 31, 2026
(Dollars in millions, except par values)	Actual	Pro Forma	Pro Forma as Adjusted
Cash and cash equivalents.............................................................	$15,852	$15,852	$
Long-term debt:
SpaceX Credit Facility (1).........................................................	$—	$—
SpaceX Bridge Loan (2)............................................................	20,000	20,000
X 2027 and X 2030 Notes........................................................	27	27
Other Financings (3)..................................................................	9,105	9,105
Unamortized deferred financing costs......................................	(21)	(21)
Total long-term debt............................................................	$29,111	$29,111
Redeemable convertible preferred stock:
Redeemable convertible preferred stock, par value $0.001; 189,155,861 shares issued and 134,451,267 shares outstanding, actual; no shares authorized, issued or outstanding, pro forma and pro forma as adjusted...............	$7,049	$—
Shareholders’ equity:
Class A common stock, par value $0.001; 2,964,501,353 shares issued and 2,882,444,444 shares outstanding, actual; 36,132,150,000 shares authorized, 6,824,581,339 shares issued and outstanding, pro forma; 36,132,150,000 shares authorized,                shares issued and outstanding, pro forma as adjusted............................................................	3	6
Class B common stock, par value $0.001; 2,421,276,530 shares issued and outstanding, actual; 6,125,000,000 shares authorized, 5,695,729,430 shares issued and outstanding, pro forma and pro forma as adjusted.................................................................................	3	6
Class C common stock, par value $0.001; 494,026,445 shares issued and outstanding, actual; 10,000,000,000 shares authorized, no shares issued or outstanding, pro forma and pro forma as adjusted..........................................	0	—
Class D common stock, par value $0.0001; no shares issued and outstanding, actual; no shares authorized, issued or outstanding, pro forma and pro forma as adjusted...............	—	—

69

Preferred stock, par value $0.001; no shares issued and outstanding, actual; 2,400,000,000 shares authorized, no shares issued or outstanding, pro forma and pro forma as adjusted.................................................................................	—	—
Additional paid-in capital.........................................................	74,083	81,126
Accumulated deficit.................................................................	(41,311)	(41,311)
Accumulated other comprehensive income.............................	1,755	1,755
Total shareholders’ equity...................................................	$34,533	$41,582
Total capitalization........................................................................	$70,693	$70,693

________________

(1)As of April 30, 2026, we had no borrowings outstanding under the SpaceX Credit Facility. In May 2026, the SpaceX Credit Facility was

amended to increase the borrowing capacity up to $5,000 million (“Amended SpaceX Credit Facility”). The Amended SpaceX Credit

Facility terminates, and all outstanding loans become due and payable, on May 19, 2031, unless the parties agree to an extension in

accordance with the terms of the Amended SpaceX Credit Facility. For more information on the SpaceX Credit Facility and Amended

SpaceX Credit Facility, please see “Management's Discussion and Analysis of Financial Condition and Results of Operations—Liquidity

and Capital Resources—Debt Agreements.”

(2)As of April 30, 2026, we had $20,000 million of borrowings outstanding under the SpaceX Bridge Loan. The SpaceX Bridge Loan matures

on September 2, 2027, subject to extension in accordance with the terms of the agreement. For more information on the SpaceX Bridge

Loan, please see “Management's Discussion and Analysis of Financial Condition and Results of Operations—Liquidity and Capital

Resources—Debt Agreements.”

(3)Includes obligations related to certain AI infrastructure assets recorded as failed sale-leaseback transactions.

70

DILUTION

Purchasers of the Class A common stock in this offering will experience immediate and substantial dilution in the

net tangible book value per share of the Class A common stock for accounting purposes. Our net tangible book value

as of March 31, 2026 was approximately $              , or $                per share of Class A common stock. Net tangible

book value per share is determined by dividing our tangible net worth (tangible assets less total liabilities) by the

total number of outstanding shares of all classes of common stock outstanding immediately prior to the completion

of this offering. After giving effect to the sale of shares of Class A common stock in this offering, the payment of

underwriting discounts and commissions and estimated offering expenses by us, the Class C Reclassification and the

Preferred Conversion as if such reclassification and conversion occurred on March 31, 2026, our adjusted pro forma

net tangible book value as of March 31, 2026 would have been approximately $                , or $                per share of

Class A common stock. This represents an immediate decrease in the net tangible book value of $                per share

of Class A common stock to Mr. Musk and other existing investors and an immediate dilution (i.e., the difference

between the offering price and the adjusted pro forma net tangible book value immediately after this offering) to

new investors purchasing shares of Class A common stock in this offering of $                per share. The following

table illustrates the per share dilution to new investors purchasing shares of Class A common stock in this offering:

Initial public offering price per share........................................................................		$
Pro forma net tangible book value per share as of March 31, 2026..........................	$
Decrease per share attributable to new investors in this offering..............................
As adjusted pro forma net tangible book value per share after giving further effect to this offering........................................................................................................
Dilution in pro forma net tangible book value per share to new investors in this offering (1)...............................................................................................................		$

_______________

(1)If the initial public offering price were to increase or decrease by $1.00 per share, then dilution in pro forma net tangible book value per

share of Class A common stock to new investors in this offering would equal $                or $                , respectively. Similarly, if the

number of shares of Class A common stock offered by us were to increase or decrease by                      shares, then dilution in pro forma net

tangible book value per share of Class A common stock to new investors in this offering would be $                or $                , respectively.

The following table summarizes, on an adjusted pro forma basis as of March 31, 2026, the total number of shares of

Class A and Class B common stock owned by Mr. Musk and other existing investors and to be owned by new

investors in this offering, the total consideration paid, and the average price per share paid by Mr. Musk and other

existing investors and to be paid by new investors in this offering at $                , calculated before deduction of

underwriting discounts and commissions and estimated offering expenses.

	Shares Acquired(1)		Total Consideration(2)		Average Price Per Share
	Number	Percent	Amount	Percent
Elon Musk and other existing investors............................................		%	$	%	$
New investors in this offering............		%	$	%	$
Total...................................................		100.0%	$	100.0%	$

______________

(1)If the underwriters exercise their option to purchase additional shares in full, Mr. Musk and other existing investors would own

approximately                % and our new investors in this offering would own approximately                % of the total number of shares of our

common stock outstanding after this offering.

(2)If the underwriters exercise their option to purchase additional shares in full, the total consideration paid by our new investors would be

approximately $                (or                %).

Each $1.00 increase or decrease in the assumed initial public offering price would increase or decrease, as

applicable, the total consideration paid by new investors and the total consideration paid by all shareholders by

$           million, assuming that the number of shares of Class A common stock offered by us remains the same and

after deducting estimated underwriting discounts and commissions and estimated offering expenses payable by us.

Similarly, an increase or decrease of                      shares in the number of shares of Class A common stock offered

by us would increase or decrease, as applicable, the total consideration paid by new investors and the total

71

consideration paid by all shareholders by $              million, assuming that the assumed initial public offering price

remains the same and after deducting estimated underwriting discounts and commissions and estimated offering

expenses payable by us.

74

MANAGEMENT’S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND RESULTS OF

OPERATIONS

The following discussion and analysis of our financial condition and results of operations should be read in

conjunction with our audited consolidated financial statements and the related notes and other financial information

included elsewhere in this prospectus. In addition to historical consolidated financial information, the following

discussion contains forward-looking statements that reflect our plans, estimates, and beliefs. Our actual results

could differ materially from those discussed in the forward-looking statements. You should review the sections titled

“Cautionary Note Regarding Forward-Looking Statements” for a discussion of forward-looking statements and

“Risk Factors” for a discussion of factors that could cause actual results to differ materially from the results

described in or implied by the forward-looking statements contained in the following discussion and analysis and

elsewhere in this prospectus. Our audited consolidated financial statements and related notes have been prepared to

reflect the retrospective combination of the companies for all periods presented as the acquisitions of xAI and X

Holdings were accounted for as transactions between entities under common control.

Our Mission

Our mission is to build the systems and technologies necessary to make life multiplanetary, to understand the true

nature of the universe, and to extend the light of consciousness to the stars. To do this, we have formed the most

ambitious, vertically integrated innovation engine on (and off) Earth with unmatched capabilities to rapidly

manufacture and launch space-based communications that connect the world, to harness the Sun to power a truth-

seeking artificial intelligence that advances scientific discovery, and ultimately to build a base on the Moon and

cities on other planets.

Starship Flight Test

Overview

Founded in 2002, SpaceX is the only company building the integrated hardware and software infrastructure of the

future across space, connectivity, and AI. At our core, we are builders. We design, manufacture, launch, and operate

products and services built on cutting-edge technologies, including the world’s most advanced rockets and

spacecraft. We safely and reliably transport astronauts, satellites, and other payloads on missions that benefit life on

75

Earth. Since 2023, we have launched more than 80% of mass to orbit for the world each year with an over 99%

mission success rate with Falcon rockets. We also operate a high-speed, low-latency global broadband data and

communications network powered by approximately 9,600 Starlink broadband and mobile satellites in Low-Earth

Orbit, delivering connectivity to millions of consumer, enterprise, and government customers across 164 countries,

territories, and other markets, as of March 31, 2026. Using our dedicated satellite-to-mobile constellation, we offer

connectivity services, supplementing terrestrial networks and substantially reducing mobile “dead zones” across

approximately 30 countries.

With the potential to improve both space exploration and life on Earth, AI accelerates SpaceX’s mission to make life

multiplanetary, to understand the true nature of the universe, and to extend the light of consciousness to the stars.

xAI, which was founded in 2023 and acquired by SpaceX in early 2026, is now an integral pillar of our vertically

integrated company. We are rapidly constructing AI compute infrastructure—starting on Earth with the goal of

extending to space—at industry-leading pace and cost efficiency. Our infrastructure supports training and inference

for Grok, which has emerged as one of the world’s most advanced frontier models. Grok is designed as a truth-

seeking AI model, built on our founder Elon Musk’s mission to enable humanity to understand the universe. We

believe that accomplishing this mission requires a truth-seeking approach to AI. We define truth seeking as the

active, relentless pursuit of what is objectively true about reality, and grounded in evidence, logic, empirical data,

and first principles thinking. Our goal is to understand and explain what the universe appears to be doing, as

accurately as current knowledge allows. Within two years of its initial model release, Grok achieved frontier-level

performance in scientific reasoning, as measured by its GPQA Diamond score, an industry benchmark that evaluates

AI models on a standardized set of questions written and validated by experts, on a faster timeline than reported by

other leading model providers. Grok also benefits from integration with X, our real-time information, entertainment,

and free speech platform, which serves as a foundational distribution and data engine for our AI ecosystem and

further enhances Grok’s truth-seeking objective.

We believe that space represents the largest economic frontier in human history, unlocking unprecedented

opportunities in orbit and on Earth. Earth has limits, so we must build infrastructure and industries in space,

expanding human capabilities to improve life on Earth and to establish life beyond. Connectivity infrastructure in

space is designed to help everyone on Earth have access to education, healthcare, entertainment, and

communications, and to enable people to overcome many traditional limits, such as physical and political borders.

We believe AI infrastructure in space can utilize the virtually limitless power of the Sun and thereby enable the use

of AI as a transformative force for understanding the universe and improving the daily lives of all humans. We

believe the convergence of these areas will enable an unprecedented expansion in the global economy, leading to an

age of abundance. Our innovations and technological advancements are redefining industries on Earth, while we aim

to create new ones on the Moon, Mars, and beyond. We are truly building the infrastructure of the future.

SpaceX is the only company that has cracked the code on accessing space at scale, revolutionizing an industry

characterized by decades of stagnation, risk aversion, and economically perverse cost structures. SpaceX upended

this paradigm through the application of first-principles thinking, which rejects industry assumptions and builds

solutions based on the fundamental laws of physics. Our intense, mission-driven, engineering-first culture and focus

on extreme vertical integration have propelled us to achieve what many deemed impossible. We have demonstrated

the ability to achieve groundbreaking technological innovations with speed, quality control, and precision. We

pioneered high-cadence, reliable, and affordable access to space with our Falcon family of rockets, with a goal to

transform the rocket launch industry into airline-like operations. In 2015, we established at least a 10-year lead over

the industry by successfully landing our first Falcon 9 booster back from space before anyone else. We have

continued to invest significantly in further increasing our lead by pursuing full and rapid reusability at scale,

including investing over $15 billion in our next-generation rocket, Starship.

We believe rocket launches and landings should be as routine and commonplace as airplanes taking off and landing.

To achieve this sort of cadence, our iterative approach emphasizes rapid designing, testing, and process

optimization, putting flight hardware in the flight environment as often as possible. This allows us to accelerate our

learning by repeatedly using and improving our systems. This has resulted in a significantly higher flight rate at

costs that are much lower than launch programs that existed before SpaceX. For example, according to NASA, the

first version of Falcon 9 in 2010 had a launch cost of approximately $2,700 per kilogram, which represented a

reduction of approximately 85% compared to the historical average launch cost per kilogram of $18,500. The first

76

version of Falcon Heavy in 2018 further reduced this cost to approximately $1,400 per kilogram, a reduction of

approximately 92% compared to the historical average cost. With the future deployment of Starship, which is

designed to be the world’s first fully and rapidly reusable spacecraft, we aim to further reduce the cost to reach orbit

by 99% or more relative to the historical average launch cost. Central to our cost advantage is the reusability of key

hardware—most notably boosters—which we recover, refurbish, and refly many times instead of discarding after

single use. This dramatically lowers per-launch costs by minimizing hardware replacement expenses and spreading

fixed production costs across repeated uses. Space flight that historically cost billions per launch now costs in the

tens of millions, fundamentally reducing the cost of space access, providing the opportunity to build new enterprises

in space.

Similarly, xAI has cracked the code in the complexities of building and scaling AI compute infrastructure, becoming

the first company to deploy a coherent gigawatt-scale AI training cluster. We believe the combination of our

proprietary AI infrastructure capability, our truth-seeking frontier model, Grok, and our access to real-time data on

X creates a formidable competitive advantage, allowing us to maintain a leading position in the development of

advanced artificial intelligence. This advantage stems from our complete vertical integration and the common vision

infused by our founder, Elon Musk. In just a few years, we have demonstrated an ability to build coherent compute

at scale and rapid speed with lower cost. COLOSSUS and COLOSSUS II collectively provide approximately 1.0

gigawatt of compute power, with additional power capacity available for data center operations. We believe speed is

a competitive advantage. In order to bring compute clusters online as fast as possible, we employ a vertically

integrated, nimble approach to construction. At COLOSSUS, we brought online the first cluster of approximately

100,000 H100 processors, approximately 130 megawatts of compute power, in just 122 days, repurposing the shell

of an existing factory. At COLOSSUS II, we brought online the first cluster of approximately 110,000 GB200

processors, approximately 210 megawatts of compute power, even faster in 91 days. As an illustrative comparison,

an industry benchmark to bring online a 100 megawatt greenfield data center is approximately two years.

Furthermore, in the case of COLOSSUS II, following the initial cluster, we brought online the second cluster of

110,000 GB300 processors and 220 megawatts of compute power in 64 days, demonstrating our ability to rapidly

scale our facilities once built. We expect that once fully operational, the next phase of expansion at COLOSSUS II

will bring online at least 220,000 additional GB300 processors and over 400 additional megawatts of compute

power. We also demonstrated asignificant improvement in cost efficiency, achieving data center construction costs

for COLOSSUS II that are considerably lower than industry benchmarks on a per megawatt basis.

We are able to deploy power and compute significantly faster than other AI companies through first-principles

thinking, behind-the-meter power generation, coupled with what we believe is the world’s largest network of

sustainable battery storage systems, and innovations in advanced liquid cooling, high-density rack layouts, and

efficient networking. Our facilities also incorporate innovative design features that limit the effects on regional

electricity pricing for neighbors and include advanced water cleaning, reclamation, and recycling processes to

support sustainable operations. We partner with utilities and communities to connect to and enhance the grid over

time, and do so while pledging to cover costs of all new power delivery infrastructure upgrades to service our data

centers, including adequate network upgrade costs, to ensure that these expenses are not passed on to the ordinary

household. Our ability to rapidly and cost-effectively scale with the latest processors keeps us ahead of competitors

who deploy traditional and more expensive methods. As a result, we believe COLOSSUS II became one of the

world’s first data centers to deploy GB200s and GB300s, the most advanced AI processors available at the time, at

significant scale, and is currently powering training for our next frontier models, including Grok-5. Furthermore,

through our Terafab initiative together with Tesla to build a manufacturing facility capable of producing 1 terawatt

per year of compute hardware, we intend to further extend our vertical integration to chip design and manufacturing

to alleviate potential future chip shortages at SpaceX, optimize compute performance, and potentially reduce overall

compute costs. Intel, which has the ability to design, fabricate, and package ultra-high-performance chips at scale,

also joined the Terafab project in early April 2026. Our shovels-to-tokens approach allows us to train and iterate our

frontier models at high velocity, accelerating development cycles, eliminating external bottlenecks, and driving

rapid, continuous improvements in model performance.

We were the first private company to develop and launch a liquid-fuel rocket to reach orbit with the successful

launch of Falcon 1 in 2008. In 2019, we were the first to begin deploying a large-scale LEO broadband satellite

77

constellation. In February 2026, we acquired xAI, the first company to build a gigawatt-scale AI training cluster and

largest coherent supercomputer. The graphic below illustrates key milestones for ourbusiness.

78

Our Repeatable Business Model

Our business model is built on a repeatable, engineering-driven framework that combines our unparalleled launch

capabilities, extreme vertical integration, rapid iteration, and disciplined capital investment to create durable, large-

scale businesses. We execute this framework through the following core principles:

1.Leverage our unparalleled launch capabilities to enable massive scale. Our rockets—with unmatched

launch cadence, best-in-class reliability, and dramatically reduced cost-to-orbit—are the foundation that we

expect will enable us to create economic opportunities in space and deliver a diversified portfolio of services.

Our launch capabilities enable large-scale deployment of assets that would not otherwise be economically

viable.

2.Identify and create new trillion-dollar market opportunities. We focus on market opportunities that are

useful for humanity and that present trillion-dollar opportunities, including global broadband and mobile

connectivity for consumers, enterprises, and governments; and AI applications and computational infrastructure.

We prioritize opportunities where structural inefficiencies or legacy technological limitations have constrained

supply.

3.Design a solution with world-class engineering and first-principles thinking. We apply physics-based

engineering and first-principles thinking to design products and systems from the ground up—boiling things

down to the most fundamental truths and reasoning up from there. This helps us drive massive, step-function

improvements in performance, scalability, and cost.

4.Apply “The Algorithm” (make less dumb, delete, optimize, accelerate, automate). We operate under a set

of core execution principles that we refer to as “The Algorithm,” a five-step iterative process that we use as our

guiding principles day-to-day. We make the requirements less dumb, delete unnecessary processes or parts

(embracing the principle that the best part is no part), only then optimize the necessary processes or parts, and

then accelerate cycle time (many entities have launched once; no one other than us has ever launched over 100

times per year), and automate only proven processes after the first four steps are completed. We apply the

Algorithm across every aspect of our organization, creating a cultural and operational standard of excellence

that has defined SpaceX since inception.

5.Vertically integrate all the way to the end customer.We design and manufacture a significant portion of our

components in-house, including engines, avionics, structures, and software, even producing the “tools that make

the tools,” enabling us to test, fail, and iterate rapidly. We can then release newer, more advanced hardware with

speed and cost efficiency.

6.Continuously drive cost down and throughput up. Through rocket reusability, manufacturing at scale,

advanced automation, and rigorous operational discipline, we continuously reduce unit costs while increasing

launch cadence, satellite network, and AI hosting capacity.

7.Generate significant cash flow and reinvest in the future. As our businesses scale, they generate significant

cash flow, which we reinvest into nascent market opportunities—driving a self-reinforcing cycle of constant

innovation and potentially creating significant additional value.

Segments in Our Vertically-Integrated Innovation Engine

We have three reportable segments in our vertically integrated innovation engine: Space, Connectivity, and AI. In

our Space segment, we design, manufacture, and launch reusable rockets to provide high cadence, reliable, and

affordable access to space at unprecedented scale. In our Connectivity segment, we operate a worldwide high-speed,

low-latency broadband data and communications network powered by over 9,600 Starlink broadband and mobile

satellites in Low-Earth Orbit, delivering connectivity to millions of consumer, enterprise, and government customers

across 164 countries, territories, and other markets. In our AI segment, we operate a highly vertically integrated AI

platform spanning our truth-seeking frontier model Grok, AI solutions for consumer and enterprise customers, X—

our real-time information, entertainment, and free speech platform—and AI computational infrastructure.

79

Our financial results reflect the strength of our operating model and our ability to create and scale multiple new

businesses:

•For the three months ended March 31, 2026, we generated revenue on a consolidated basis of $4,694 million,

loss from operations of $(1,943) million and Adjusted EBITDA of $1,127 million. In 2025, we generated

revenue on a consolidated basis of $18,674 million, loss from operations of $(2,589) million and Adjusted

EBITDA of $6,584 million. Our Space and Connectivity segments contributed the substantial majority of our

consolidated revenue in the three months ended March 31, 2026 and the year ended December 31, 2025,

demonstrating the benefits of their scale and operating leverage in our vertically integrated business model;

•For the three months ended March 31, 2026, our Space segment generated revenue of $619 million, loss from

operations of $(662) million, and Segment Adjusted EBITDA of $(351) million. In 2025, our Space segment

generated revenue of $4,086 million, loss from operations of $(657) million, and Segment Adjusted EBITDA of

$653 million. Additionally, our Space segment funded $930 million and $3,004 million in research and

development expense during the three months ended March 31, 2026 and the year ended December 31, 2025,

respectively, for our next-generation Starship launch vehicle program. Starship is designed to enable a step-

function change in our launch capability across reusability, payload capacity, and launch cadence, and is the key

enabler of our long-term growth strategy by unlocking entirely new categories of missions;

•For the three months ended March 31, 2026, our Connectivity segment generated revenue of $3,257 million,

income from operations of $1,188 million, and Segment Adjusted EBITDA of $2,087 million. Our Connectivity

segment, primarily driven by Starlink, generated revenue of $11,387 million, income from operations of $4,423

million, and Segment Adjusted EBITDA of $7,168 million in 2025, representing year-over-year growth of

49.8%, 120.4%, and 86.2%, respectively, benefiting from subscriber growth, increasing enterprise adoption, and

continued improvement in network efficiency;

•In our newly acquired AI segment, we plan to prioritize growth and investment to capture significant

opportunities in AI applications and compute infrastructure. For the three months ended March 31, 2026, our AI

segment generated revenue of $818 million, loss from operations of $(2,469) million, and Segment Adjusted

EBITDA of $(609) million. In 2025, our AI segment generated revenue of $3,201 million, loss from operations

of $(6,355) million, and Segment Adjusted EBITDA of $(1,237) million, reflecting its earlier stage of

development and continued investments to support long-term growth opportunities in AI; and

•For the three months ended March 31, 2026, capital expenditures for our Space segment was $1,052 million, for

our Connectivity segment was $1,332 million and for our AI segment was $7,723 million. In 2025, capital

expenditures for our Space segment was $3,832 million, for our Connectivity segment was $4,178 million and

for our AI segment was $12,727 million. 

Segment Adjusted EBITDA is a non-GAAP measure. Please refer to the section titled “—Non-GAAP Financial

Measures” for additional information on our non-GAAP financial measures, including reconciliations of Segment

Adjusted EBITDA to segment income (loss) from operations, the most directly comparable GAAP measure.

Space

Since our founding in 2002, SpaceX has cracked the code on accessing space at scale, transforming an industry

characterized by decades of stagnation, risk aversion, and economically perverse cost structures. We design,

manufacture, launch, and refurbish reusable launch vehicles that provide cost-efficient, reliable, and high-cadence

access to space for our own purposes as well as for third-party commercial and government customers. In 2025, we

launched from four primary launch pads in the United States and successfully recovered boosters across seven

landing facilities including autonomous drone ships and catch towers based on the vehicle type and mission profile.

Our extensive vertical integration and end-to-end control over the entire value chain, from design to launch to

operations, allows us to achieve unprecedented speed and cost efficiency.

80

Falcon 9 First Stage Booster Landing

As of March 31, 2026, SpaceX had launched a total mass to orbit of approximately 7,400 metric tons with an over

99% mission success rate across our Falcon rockets. We have completed approximately 650 orbital space launches,

and over 540 of those launches were completed by a flight-proven Falcon rocket. In 2025 alone, SpaceX completed

170 missions across Falcon and Starship vehicles and 159 flight-proven booster launches with an over 99% success

rate on attempted booster recoveries. We launched over 2,200 metric tons, representing over 80% of mass to orbit

for the world in 2025. With the first successful launch of Falcon 1 in 2008, we became the first private company to

successfully launch a liquid-fueled rocket to Earth’s orbit. Just two years later, in 2010, the commercial debut of the

Falcon 9 rocket revolutionized space access by delivering unprecedented cost efficiency. For example, according to

NASA, the first version of Falcon 9 in 2010 reduced launch cost to approximately $2,700 per kilogram, which

represented a reduction of approximately 85% compared to the historical average launch cost per kilogram of

$18,500. The first version of Falcon Heavy in 2018 further reduced this cost to $1,400 per kilogram, a reduction of

approximately 92% compared to the historical average. We have also reduced our internal cost of launch through a

combination of engineering improvements, manufacturing efficiencies, and economies of scale—most notably,

through our ability to drive more frequent reuse of rockets.

We generate Space revenue primarily through launch and mission services of Falcon 9, Falcon Heavy, and Dragon

provided to commercial and government customers. We fly to LEO, MEO, GEO, lunar, and interplanetary

trajectories, as well as the International Space Station. Our Space segment revenue is derived from fixed-price

contracts related to the development and provision of launch services for both commercial customers and

governmental agency space programs, either at a “point in time” or “over time.”

We manage our Space segment to support our businesses and those of our customers. We plan launches and allocate

payloads in advance, although it can be difficult to manage the timing of customer payload arrivals. When an

expected customer payload for a planned launch is not available, we instead use launch capacity for our satellites. As

a result, we adjust expected launch payloads frequently, impacting period-to-period financial comparison. For a

majority of customer payloads, revenue and costs are primarily recognized at the launch or deployment of the

customer’s spacecraft to its intended orbit, with some revenues and costs being recognized over time. For launches

dedicated to deploying our Starlink satellites, we capitalize the associated costs within our Connectivity segment and

depreciate them over time, and we do not recognize revenue for those launches in our Space segment. We allocate a

81

significant amount of launch capacity to our Connectivity segment, and expect to allocate a significant amount to

our AI segment in the future. Our Space segment revenue only reflects customer launches and other customer

activities. As a result, notwithstanding an increasing launch cadence, our Space segment has relatively lower

revenue scale and revenue growth compared to our other segments, though its financial results do not reflect the

foundational strategic value that it provides to us in bolstering the growth of our Connectivity and AI segments.

Connectivity. Starlink provides global access to high-speed internet, including underserved rural and remote

communities worldwide. As of March 31, 2026, we had approximately 9,600 Starlink broadband and mobile

satellites in Low-Earth Orbit, providing broadband connectivity to approximately 10.3 million Starlink Subscribers

across 164 countries, territories, and other markets. We also provide satellite-to-mobile texting and over-the-top

voice services to approximately 7.4 million monthly unique devices across approximately 30 countries.

Starlink Mini

•Starlink Consumer Broadband.We operate the world’s largest and most advanced space-based internet

broadband service with median latency at approximately 25 milliseconds as of March 31, 2026. We provide

fiber-like download speeds—at a median of 225 Mbps during peak hours for residential users as of March 31,

2026—and the technological capability to provide service everywhere on Earth, including the poles. This

service quality is enabled by our vast network of approximately 9,600 Starlink broadband and mobile satellites

in Low-Earth Orbit, which accounted for approximately 75% of all active maneuverable satellites in orbit as of

March 31, 2026. We expect to commence deploying our next-generation V3 satellites, designed to offer one

Tbps of downlink capacity per satellite, using Starship in the second half of 2026. We expect that a single

Starship launch will be capable of deploying up to 60 V3 satellites to LEO, representing a potential twenty-fold

increase in Starlink downlink capacity deployed relative to a Falcon 9 launch. As of March 31, 2026, we had

approximately 10.3 million Starlink Subscribers, up approximately 105% from 5.0 million subscribers a year

prior. We charge our Starlink Subscribers a monthly subscription fee, which varies based on geographic market

and download speed, plus typically a one-time upfront terminal cost.

•Enterprise Solutions. SpaceX is a critical partner to a wide array of enterprises. We offer Starlink’s high-

speed, low-latency, reliable internet services to enterprise customers across industries including construction,

agriculture, retail, telecom, hospitality, aviation, maritime, and land mobility. Starlink’s unique capabilities are

well‑suited for deployments across field offices, remote worksites, research stations, drilling rigs, rural

82

hospitals, aircraft, cruise ships, trains, and hotels. Our enterprise customers include companies such as United

Airlines, Carnival, Maersk, and John Deere, among others. We also serve a broad fixed‑site customer base

across industries such as retail and financial services that require high availability for critical operations as well

as reliable connectivity in remote or hard-to-serve locations. As companies continue to invest in secure and

resilient networks and backup systems to keep critical infrastructure online—such as point‑of‑sale and payment

processing systems—we often start as a backup solution and then transition to being the primary solution. Our

enterprise contracts are based on a combination of subscriptions, data consumption, capacity, or other pricing

models depending on each customer’s particular needs. Since 2023, no Starlink Enterprise customer having

contributed more than $750,000 of annual revenue has voluntarily discontinued their service, demonstrating the

strong performance and value of our offering. This is despite the ability of our customers to cancel the service at

any time.

•Government Solutions. For our government customers, we provide high-speed, resilient connectivity for

public services, social impact, humanitarian efforts, and disaster response in even the most remote and

challenging environments. Examples include support for the FEMA in coordinating disaster recovery after

hurricanes and wildfires, the NOAA for at-sea testing and environmental monitoring, the Government of the

Philippines for linking remote islands, schools, and public institutions, the Government of Jamaica for

improving digital access in remote and maritime areas, and the Government of Ecuador for supporting

education and healthcare connectivity in isolated communities. Separately with Starshield, we have leveraged

our commercial LEO satellite constellation engineering learnings and operational experiences to develop a

secure, dedicated satellite network designed specifically for United States Government customers and national

security applications.

•Starlink Mobile. We provide satellite-to-mobile connectivity, supplementing terrestrial networks and

substantially reducing mobile “dead zones” across approximately 30 countries. We partner with MNOs

including major wireless carriers like T-Mobile in the United States, and other international operators including

One NZ, Optus, Telstra, Rogers, KDDI, Salt, Entel, Kyivstar, and VMO2. Through these partnerships, we

enable consumers, businesses, and public-sector customers to use their existing phones in more places, support

critical connectivity during disasters and power outages, and open new applications for low-bandwidth mobile

and IoT devices. Our current capabilities under our “V1” constellation (consisting of approximately 650 V1

Mobile satellites in orbit) include light data, text messaging (SMS), and over-the-top voice services (e.g.,

WhatsApp and FaceTime). We are developing more comprehensive satellite-to-mobile services, including

broadband data and IoT connectivity, which are expected to deliver resilient, infrastructure-independent

connectivity worldwide and enable 5G connectivity. We have partnerships with approximately 30 MNOs on six

continents, covering an area that is home to approximately 1.9 billion people. We charge MNOs either a fixed

fee or a per-mobile user fee-based amount, which is typically passed through to the customer via the carrier as

an “add-on” feature.

We generate revenue in our Connectivity segment primarily through subscription fees from consumer subscribers.

We drive consumer revenue through monthly subscription fees based on geographic market and download speed,

recognizing revenue ratably over the service period, plus typically a one-time sale of a kit. In addition, we generate

revenue from enterprises through contracts structured as a combination of subscriptions, data consumption, and

capacity, or on a percentage-of-completion basis, depending on each customer’s particular needs. We generate

government revenue via long term contracts for Starshield, a secure satellite network designed specifically for

government customers and national security applications. We also earn Starlink Mobile revenue through revenue-

sharing arrangements with MNO partners, based on connectivity services included in their plans.

In 2025, revenue from consumer subscribers represented over 60% of Connectivity segment revenue. We expect 

revenue from consumer subscribers, as well as enterprise and government customers, to be  the primary driver of

Connectivity segment growth, and that Starlink Mobile will become a significant new contributor of Connectivity

segment revenue.

AI. We operate a highly vertically integrated AI platform spanning gigawatt-scale AI compute infrastructure, our

truth-seeking frontier AI model, Grok, AI solutions for consumer and enterprise customers, and X, our real-time

information, entertainment, and free speech platform. We believe AI is rapidly converging toward AGI, where

83

human cognitive capabilities can be replicated and scaled at machine speeds, profoundly augmenting human

productivity. Once an AGI system exists, its true value derives from the ability to create limitless duplicates of

human-like intelligence, necessitating vast computational resources and cost-efficient deployment to achieve

meaningful scale. Without large-scale, power-efficient infrastructure, AGI cannot be deployed broadly or

economically—making such infrastructure a critical strategic differentiator.

COLOSSUS II Facility in Memphis, Tennessee

•AI Compute Infrastructure. xAI has established a leading position in building and scaling terrestrial AI

compute infrastructure, becoming the first company to deploy a coherent gigawatt-scale AI training cluster. Our

AI compute facilities, COLOSSUS and COLOSSUS II, collectively provide approximately 1.0 gigawatt of

compute power, with additional power capacity available for data center operations. Our first-principles

thinking enables us to build coherent compute at scale and at rapid speed with lower costs than most other

companies in the industry. We brought the first cluster of COLOSSUS online in 122 days, repurposing the shell

of an existing factory, and the first cluster of COLOSSUS II online even faster in 91 days. As an illustrative

comparison, an industry benchmark to bring online a 100 megawatt greenfield data center is approximately two

years. We also demonstrated a significant improvement in cost efficiency, achieving data center construction

costs for COLOSSUS II that are considerably lower than industry benchmarks on a per megawatt basis. This

dual speed and cost advantage stems from our complete vertical integration and the shared culture infused by

our founder, Mr. Musk, across our Space, Connectivity, and AI segments. The addition of Terafab, an initiative

together with Tesla to build a manufacturing facility capable of producing 1 terawatt per year of compute

hardware, aims to further extend our vertical integration to chip design and manufacturing to alleviate potential

future chip shortages at SpaceX, optimize compute performance, and potentially reduce overall compute costs.

Intel, which has the ability to design, fabricate, and package ultra-high-performance chips at scale, has also

joined the Terafab project. We believe that the key constraints in the continued growth of AI are physical—chip

manufacturing, data center infrastructure, and power generation; the future of AI will be determined by the

control of the physical stack.

•Truth-Seeking Frontier Model. xAI has developed one of the world’s most advanced, truth-seeking frontier

models with Grok. Since launching Grok-1 in November 2023, we have released four major versions and

notable variations thereof, achieving one of the fastest iteration cycles in the industry, culminating in Grok-4.3

84

(April 2026). Building on this trajectory, we expect to continue scaling Grok through subsequent generations.

Ongoing training of next‑generation models is expected to scale toward multiple trillions of parameters, which

could represent a step change in reasoning in depth and overall intelligence. In this context, the number of

parameters refers to the scale of the model, where parameters are the internal numerical values, such as

“weights,” that are adjusted during training to enable the model to recognize patterns and relationships in data.

A larger number of parameters generally allows the model to capture more complex relationships, store greater

amounts of knowledge, and achieve higher levels of reasoning capability.Within two years of its initial model

release, Grok achieved frontier-level performance in scientific reasoning, as measured by its GPQA Diamond

score, an industry benchmark that evaluates AI models on a standardized set of questions written and validated

by experts, on a faster timeline than reported by other leading model providers. This accelerated rate of

innovation stems from our highly vertically integrated stack: full ownership of training infrastructure, access to

the world’s most powerful compute clusters, and relentless focus on truth seeking and real-world utility. A key

competitive differentiator is Grok’s deep integration with X, enabling proprietary access to a real-time

information stream of approximately 350 million daily posts, which enhances freshness, relevance, and

contextual awareness for Grok. This direct, real-time access to the information and human discourse on X

enhances Grok’s truth-seeking capabilities by grounding outputs in up-to-date knowledge and diverse

viewpoints. We believe that this combination of compute infrastructure scale and the massive dataset available

to us through X, subject to some limitations for certain content, has allowed us to achieve industry-leading

performance and provide model outputs that analyze real-time information on global events. We expect that our

compute infrastructure and direct access to real-time data via X constitute substantial performance advantages

for Grok that will result in increasingly rapid and dramatic iteration cycles.

•Consumer and Enterprise Applications. We leverage our leading frontier models and compute infrastructure

to deliver consumer and enterprise applications. In under six months, we developed Grok Voice, a real-time

speech engine, including in multilingual performance. Our image and video generation system, Imagine,

produced approximately 10 billion images and over 2 billion videos per month, on average, for the quarter

ending March 31, 2026. Together with Tesla, we are also developing Macrohard, an agentic AI platform

designed to be capable of fully emulating digital workflows and augmenting human operation of computers—

from coding and product development to management and entire business processes—using sophisticated

autonomous agents. We believe Macrohard will have the potential to fundamentally transform how companies

are structured and operate, thereby allowing dramatic increases in human productivity. In addition, we believe

our existing government relationships and track record as large government contractors are a structural

advantage as governments become significant consumers of AI applications.

Our integrated AI platforms across Grok and X have over 1.3 billion supported accounts active in the last

twelve months ended March 31, 2026, including approximately 550 million MAUs, up from over 1.1 billion

supported accounts and approximately 520 million MAUs as of December 31, 2025. Of our MAUs, we had

approximately 117 million MAUs that used Grok’s AI features as of March 31, 2026. While MAUs provide an

estimated measure of the size and engagement of our user base, we are focused on revenue and operating

margin, and manage our business with the objective of driving sustainable revenue growth and profitability

rather than with the primary objective of growing or maintaining MAU levels.

We also monetize user activity through high-impact advertising inventory on X. We believe X’s scale, real-time

engagement, and integration with Grok provide a differentiated foundation for building a unified user

experience across communication, content discovery, commerce, and financial services, among others. For

enterprises that advertise on X, we offer large-scale user engagement, real-time content, and advanced AI-

driven performance marketing tools. For enterprises, we offer tailored deployments of Grok customized to

specific workflows and security needs through Grok Business and Grok Enterprise, sold on license-,

consumption-, or outcome-based pricing models.

Our Capital Allocation and Funding Strategy

Since our beginning, we have managed through multiple investment cycles. We initially raised capital to fund what

is now our Space segment, which generates revenue from commercial and government customers while serving as

the backbone for our Connectivity segment. We invested in our Connectivity segment as we generated Segment

85

Adjusted EBITDA from our Space segment, along with additional equity capital that we raised externally, creating a

segment that generates predictable and recurring revenue from consumer, enterprise, and government customers. We

continue to invest meaningfully in both our Space and Connectivity segments to build out the infrastructure of the

future through our next-generation Starship launch platform and our expanded Starlink broadband and mobility

networks.

We have a stellar track record of capital allocation and value creation in Space and Connectivity. Since SpaceX’s

founding in 2002, we have raised over $9 billion of equity capital to fund the development and growth of these two

business segments. The Space segment became Segment Adjusted EBITDA positive on a sustained basis beginning

in 2018 and the Connectivity segment became in aggregate Segment Adjusted EBITDA positive on a sustained basis

beginning in 2023. In 2025, our Space segment generated a loss from operations of $(657) million and Segment

Adjusted EBITDA of $653 million, including the impact of funding $3,004 million in research and development

expense for our next-generation Starship launch vehicle program. In 2025, our Connectivity segment generated

income from operations of $4,423 million and Segment Adjusted EBITDA of $7,168 million.

We acquired xAI in February 2026, which forms the basis of our AI segment. We expect to allocate substantial

capital to expand our compute infrastructure, and we expect a multi-year investment horizon before these

deployments translate into sustained positive AI Segment Adjusted EBITDA. During this investment period, our

capital expenditures will scale as quickly as we are able to deploy power and compute to address the $26.5 trillion

potential market opportunity for AI. We plan to access a range of debt and equity financing solutions available to us

as a public company to fund future investments in growth and to maintain strong liquidity. We aim to maintain an

investment grade credit rating.

Segment Adjusted EBITDA is a non-GAAP measure. Please refer to the section titled “—Non-GAAP Financial

Measures” for additional information on our non-GAAP financial measures, including reconciliations of Segment

Adjusted EBITDA to segment income (loss) from operations, the most directly comparable GAAP measure.

Key Business Metrics

We use the following key business metrics to evaluate our business, measure our performance, identify trends,

formulate business plans, and make strategic decisions.

Space

In our Space segment, we use mass to orbit and launches as key business metrics to measure our scale and

throughput. Mass to orbit and launches grow more rapidly than Space segment revenue because these metrics

include our internal constellation deployments from which we do not recognize inter-segment revenue.

Mass to Orbit: Mass to orbit is the total kilograms of payload that we deploy to orbit in a given period, and is a key

indicator of SpaceX’s capacity and scalability that supports Space revenue and drives expansion across our

Connectivity and AI segments. We calculate this metric by summing verified mass, including Starlink satellites,

customer payloads, and development cargo, from all successful orbital and flight tests. This measure excludes failed

or scrubbed attempts. We increased mass to orbit from 1,210 metric tons in 2023 to 1,699 metric tons in 2024 to

2,213 metric tons in 2025, and from 450 metric tons in the three months ended March 31, 2025 to 556 metric tons in

the three months ended March 31, 2026. In 2023, 2024, and 2025, mass to orbit included 205, 282, and 312 metric

tons attributable to customer payloads, respectively, and 1,005, 1,418, and 1,901 metric tons attributable to internal

payloads, respectively (the amounts presented may not add up to the corresponding totals due to rounding). Falcon 9

launches contribute steadily at an average capacity of 13 metric tons per mission since 2023 to various orbits while

we transition to Starship. As the most powerful launch system ever developed, we expect that Starship V3 will be

able to carry a payload of 100 metric tons, with future generations of Starship being designed to double this payload.

86

Launches: Launches are a key measure of our operational scale, which in turn supports our revenue growth and

mission to expand humanity’s presence in space. Launches in a period represent the sum of all successful orbital and

flight tests across our rockets, including internal Starlink deployments, development tests, and launches for our

third-party customers, and excluding any cancellations or scrubs that occurred in that period. Falcon 9 is the most

active orbital launch vehicle today, with approximately 620 orbital space launches as of March 31, 2026, and an

over 99% mission success rate. During the three months ended March 31, 2026, we launched 40 Falcon rockets, of

which 39 were flight-proven booster launches, and in 2025, we launched 165 Falcon 9 rockets, of which 157 were

flight-proven booster launches. While we have steadily increased our Falcon 9 launch cadence over recent years, we

expect Falcon 9 launches to decrease over time. While Falcon 9 currently drives the majority of our launch activity,

we expect Starship, which is designed to be the world’s first fully, rapidly, reusable launch vehicle, to become a

larger contributor to our launch volume as it enters operational service. To date, we have executed 11 Starship flight

tests to advance our goal of rapidly and fully reusable orbital capability, a breakthrough we believe will transform

our launch economics and benefit both our business and customers who rely on our launch services. We have also

scheduled a 12th flight test, which will debut the next generation Starship vehicle and Super Heavy booster,

powered by the next evolution of our Raptor engine and launching from a newly designed pad at Starbase.We

allocate a significant amount of launch capacity to our Connectivity segment, and expect to allocate a significant

87

amount to our AI segment in the future. Our Space segment revenue only reflects our customer launches and

customer activities.

__________________

(1)With respect to Falcon launches, the number of launches for the years ended December 31, 2023, 2024, and 2025 totaled 96, 134, and 165,

respectively, of which customer launches totaled 33, 45, and 43, respectively, and internal launches totaled 63, 89, and 122, respectively.

The number of Falcon launches for the three months ended March 31, 2025 and 2026 totaled 36 and 40, respectively, of which customer

launches totaled 12 and 7, respectively, and internal launches totaled 24 and 33, respectively. We designate a launch as a “customer launch”

if an external customer payload constitutes the primary payload (i.e., where the principal objective is to deliver the customer payload) and

the mission parameters (e.g., launch window, orbital parameters, mission profile) are designed around the primary payload’s requirements.

To date, all Starship launches have been classified as internal.

Connectivity

In our Connectivity segment, we view Starlink Subscribers and Starlink Subscriber ARPU as key business metrics to

evaluate our growth and monetization.

Starlink Subscribers: We define a Starlink Subscriber as a unique Service Line that is directly assigned to a

Starlink.com account registered to a person or entity that does not have a direct, negotiated agreement with the

Starlink sales team. A Service Line is an individual instance of Starlink broadband internet service provisioned

under a subscription plan, generally associated with a specific Starlink terminal or group of terminals, and billed

according to Starlink’s service plans and terms of service. The number of Service Lines is distinct from the number

of unique devices, account holders, end users or physical persons. An individual, household, or business may share a

single Service Line among multiple end-users. Likewise, an individual, household, or business may maintain

multiple service lines (e.g., both a Residential Service Line and a separate Roam Service Line, which would be

defined as two separate Service Lines and therefore two Starlink Subscribers).

We use this measure to assess the adoption of Starlink as we expand within and across geographies and business

segments. Starlink Subscribers includes both Personal (e.g., Residential and Roam) and Business (e.g., Local

Priority and Global Priority) subscription plans, but does not include managed enterprise and government customers

with contracts in domains including aviation, maritime, land mobility, fixed sites and government entities. We

calculate Starlink Subscribers for a period as the number of unique Service Lines at the end of the period. Starlink

88

Subscribers totaled approximately 10.3 million and 5.0 million, up 105% and 91% on a year-over-year basis, in the

quarters ended March 31, 2026 and March 31, 2025, respectively.

Starlink Subscriber ARPU: We calculate ARPU as service revenue generated from Starlink Subscribers during the

period divided by (i) the average number of Starlink Subscribers during the period and by (ii) the number of months

in the period. Our strategy is focused on driving sustainable revenue growth and expanding our margins through

operational efficiencies and technological advancements, rather than prioritizing increases in ARPU. This approach

aligns with our long-term vision of expanding global connectivity and market access. We generally expect Starlink

Subscriber ARPU to continue to decline over the next few years as the portion of our subscriber base outside North

America continues to grow, as we add lower priced service plans, and as we adjust the monthly service plan fees we

charge for broadband offerings. However, we expect these dynamics to be offset by increased scale and

technological advancement in our launch, satellite, and user terminal operations, ultimately supporting overall

revenue growth and cost reduction. Our Starlink Subscriber monthly ARPU decreased from $86 per month for the

three months ended March 31, 2025 to $66 per month for the three months ended March 31, 2026 and from $91 per

89

month in 2024 to $81 per month in 2025. These decreases were driven primarily by international expansion and the

addition of lower priced service plans.

AI

Nameplate Compute Draw: We calculate Nameplate Compute Draw for a period as the number of GPUs installed in

our data centers at the end of the period multiplied by their respective all-in power draw. Nameplate Compute Draw

reflects installed capacity and does not represent actual power consumption or utilization. It does not include power

we install and use for our supporting infrastructure such as cooling systems, power distribution losses, lighting,

security systems, or facility-level overhead. Our Nameplate Compute Draw increased to 1.0 gigawatt as of March

90

31, 2026 as we brought COLOSSUS and COLOSSUS II online. We use this metric to assess our ability to deploy

and scale compute capacity.

Segment Income (Loss) from Operations

Space Income (Loss) from Operations

Space loss from operations for the three months ended March 31, 2026 increased by $592 millionto $(662) million

compared to $(70) million for the three months ended March 31, 2025, primarily driven by an accelerated

investment in development of the Starship vehicle as well as launch facilities to support future Starship launches,

and a decrease in revenue from customer launches, partially offset by a decrease in cost of revenue, selling, general,

and administrative expenses and impairment.

Space income (loss) from operations for the year ended December 31, 2025decreased by $678 millionto $(657)

millioncompared to $21 million for the year ended December 31, 2024, while Space income (loss) from operations

for the year ended December 31, 2024 increased by $22 million to $21 million for the year ended December 31,

2024 compared to $(1) million for the year ended December 31, 2023. The year-over-year decrease in 2025 was

primarily driven by an accelerated investment in development of the Starship vehicle as well as launch facilities to

support future Starship launches, partially offset by an increase in revenue and decrease in cost of revenue.

Connectivity Income (Loss) from Operations

Connectivity income from operations for the three months ended March 31, 2026 increased by $155 millionto

$1,188 millioncompared to $1,033 million for the three months ended March 31, 2025, primarily driven by

increased revenue from our consumer subscribers (composed of 104.7% growth in Starlink Subscribers, offset by a

22.9%decline in Starlink Subscriber ARPU, primarily due to international expansion and the addition of lower

priced service plans) and enterprise business, partially offset by higher depreciation of capitalized launch and

satellite costs due to the increase in Starlink flights, as well as higher operating expenses including ground operating

costs and international expansion costs to support and drive subscriber growth.

Connectivity income from operations for 2025 increased by $2,417 million to $4,423 million compared to $2,006

million for the year ended December 31, 2024 while Connectivity income from operations for the year ended

91

December 31, 2024increased by $1,537 million to $2,006 million compared to $469 million for the year ended

December 31, 2023. The year-over-year increase in 2025 was primarily driven by increased revenue from growth of

our consumer and enterprise customers by $2,378 million and $1,410 million, respectively, partially offset by higher

depreciation of capitalized launch and satellite costs due to the increase in Starlink flights, as well as higher

marketing and international expansion costs to drive subscriber growth.

AI Income (Loss) from Operations

AI loss from operations for the three months ended March 31, 2026 increased by $1,533 millionto $(2,469) million

compared to $(936) million for the three months ended March 31, 2025, primarily driven by higher cloud computing

and GPU depreciation costs, data center infrastructure and employee expenses, partially offset by higher revenue.

AI loss from operations for 2025 increased by $4,794 millionto $(6,355) million compared to $(1,561) million for

the year ended December 31, 2024, while AI loss from operations for the year ended December 31, 2024 decreased

by $2,412 million to $(1,561) million compared to $(3,973) million for the year ended December 31, 2023. The

increase in 2025 was primarily driven by higher cloud computing costs, facilities-related costs and employee

expenses, partially offset by higher revenue.

Segment Adjusted EBITDA

Segment Adjusted EBITDA is defined as segment income (loss) from operations excluding (i) depreciation and

amortization, (ii) share-based compensation, (iii) restructuring charges and (iv) impairment.

Space Segment Adjusted EBITDA

Space Segment Adjusted EBITDA for the three months ended March 31, 2026 decreased by $575 millionto $(351)

millioncompared to $224 million for the three months ended March 31, 2025, primarily driven by an accelerated

investment in development of the Starship vehicle as well as launch facilities to support future Starship launches,

and a decrease in revenue from customer launches, partially offset by a decrease in cost of revenue, selling, general,

and administrative expenses.

Space Segment Adjusted EBITDA for 2025 decreased by $501 million to $653 million compared to $1,154 million

in 2024, while Space Segment Adjusted EBITDA for 2024 increased by $157 million to $1,154 million compared to

$997 million in 2023. The year-over-year decrease in 2025 was primarily driven by an accelerated investment in

development of the Starship vehicle, as well as launch facilities to support future Starship launches, partially offset

by an increase in NASA Cargo Resupply Services (CRS) for additional missions to the International Space Station,

along with increased revenue from a U.S. Department of War contract. Our Space Segment Adjusted EBITDA is

also driven by the reusability and efficiency of our rockets, which boosts cadence and reliability and supports a

diversified base of commercial and government customers. These efforts have created a strong foundation for our

Space Segment Adjusted EBITDA, and we believe position us to unlock further high-value opportunities in the

expanding space economy.

Connectivity Segment Adjusted EBITDA

Connectivity Segment Adjusted EBITDA for the three months ended March 31, 2026 increased by $469 millionto

$2,087 millioncompared to $1,618 million for the three months ended March 31, 2025, primarily driven by higher

revenue from growth in consumer and enterprise revenue. Consumer revenue was composed of 104.7% growth in

Starlink Subscribers, offset by a 22.9%decline in Starlink Subscriber ARPU, primarily due to international

expansion and the addition of lower priced service plans. Enterprise and government revenue had an increase

primarily driven by the growth in our aviation, maritime, mobility, and other enterprise business, partially offset by a

decrease in our government business. These increases in revenue were offset by higher operating expenses for

international expansion, and higher research and development costs.

Connectivity Segment Adjusted EBITDA for 2025 increased by $3,319 million to $7,168 million compared to

$3,849 million in 2024 while Connectivity Segment Adjusted EBITDA for 2024 increased by $2,247 million to

$3,849 million compared to $1,602 million in 2023. The year-over-year increase in 2025 was primarily driven by

92

higher revenue from growth in our consumer and enterprise customers, partially offset by higher marketing and

international expansion costs to grow our subscribers, as well as higher research and development costs for our next-

generation product development. We have driven our strong sequential Connectivity Segment Adjusted EBITDA

growth by expanding the scale and efficiency of our LEO satellite constellations and our highly verticalized supply

chain, which has delivered major cost reductions in user terminal production.

AI Segment Adjusted EBITDA

AI Segment Adjusted EBITDA for the three months ended March 31, 2026 decreased by $497 millionto $(609)

millioncompared to $(112) million for the three months ended March 31, 2025, primarily driven by higher cloud

compute and data center infrastructure and operating costs, and employee compensation expenses, partially offset by

higher revenue.

AI Segment Adjusted EBITDA for 2025 decreased by $1,584 million to $(1,237) million compared to $347 million

in 2024 while AI Segment Adjusted EBITDA for 2024 decreased by $875 million to $347 million, compared to

$1,222 million in 2023. The decrease in 2025 was primarily driven by higher cloud computing costs, facilities-

related costs and employee expenses, partially offset by higher revenue. AI Segment Adjusted EBITDA is primarily

driven by our strategy to rapidly and cost-effectively scale compute infrastructure. We expect to continue to expand

our terrestrial data centers, and to launch orbital data centers, and we expect a multi-year investment horizon before

these deployments translate into sustained positive Segment Adjusted EBITDA for our AI segment.

Segment Adjusted EBITDA is a non-GAAP measure. Please refer to the section titled “—Non-GAAP Financial

Measures” for additional information on our non-GAAP financial measures, including reconciliations of Segment

Adjusted EBITDA to segment income (loss) from operations, the most directly comparable GAAP measure.

Capital Expenditures

The following table presents our capital expenditures by segment:

	Three Months Ended March 31,		Year Ended December 31,
(in millions)	2026	2025	2025	2024	2023
Space..................................................	$1,052	$759	$3,832	$2,032	$1,497
Connectivity.......................................	1,332	814	4,178	3,498	2,455
AI........................................................	7,723	2,567	12,727	5,633	463
Total Capital Expenditures.................	$10,107	$4,140	$20,737	$11,163	$4,415

Space Capital Expenditures

Space capital expenditures for the three months ended March 31, 2026 increased $293 million to $1,052 million

compared to $759 million for the three months ended March 31, 2025. The increase was primarily driven by

increased investment in our launch site infrastructure for Starship.

Space capital expenditures for 2025 increased $1,800 million to $3,832 million compared to $2,032 million in 2024,

while Space capital expenditures for 2024 increased $535 million to $2,032 million compared to $1,497 million in

2023. The increase in each year-over-year period was primarily driven by increased investment in our launch site

infrastructure for Starship.

Connectivity Capital Expenditures

Connectivity capital expenditures for the three months ended March 31, 2026 increased $518 million to $1,332

million compared to $814 million for the three months ended March 31, 2025. The increase was primarily driven by

higher satellite and ground equipment costs as we continue to increase our number of satellites and grow our satellite

network.

Connectivity capital expenditures for 2025 increased $680 million to $4,178 million compared to $3,498 million in

2024, while Connectivity capital expenditures for 2024 increased $1,043 million to $3,498 million compared to

93

$2,455 million in 2023. The increase in each year-over-year period was primarily driven by higher satellite and

ground equipment costs as we continue to increase our number of satellites and grow our satellite network. 

AI Capital Expenditures

AI capital expenditures for the three months ended March 31, 2026 increased $5,156 million to $7,723 million

compared to $2,567 million for the three months ended March 31, 2025. The increase was primarily driven by

investments in the rapid expansion of our terrestrial data centers, including the development, construction, and

equipping of new facilities and supporting infrastructure.

AI capital expenditures for 2025 increased $7,094 million to $12,727 million compared to $5,633 million in 2024,

while AI capital expenditures for 2024 increased $5,170 million to $5,633 million compared to $463 million in

2023. This increase was primarily driven by significant investments in the rapid expansion of our terrestrial data

centers, including the development, construction, and equipping of new facilities and supporting infrastructure.

Drivers of Our Performance

Developing Starship. Starship is our next-generation vehicle that we expect will dramatically expand our launch

capability through full and rapid reusability combined with unprecedented mass to orbit capability. As the most

powerful launch system ever developed, we expect that Starship V3 will be able to carry a payload of 100 metric

tons, and that future generations could reach 200 metric tons, potentially as soon as Starship V4. Starship is central

to our goal of unlocking growth through our unique vertically integrated business model. Starship is expected to be

the only vehicle with fully reusable first and second stages, which is critical to reducing launch costs and increasing

launch cadence. We believe that Starship can eventually reduce the cost to reach orbit by 99% or more relative to the

historical average launch cost per kilogram according to NASA of $18,500, establishing a scalable path to creating

the infrastructure of the future, such as orbital AI compute.

We have already demonstrated catching and reusing the first stage booster for Starship through our innovative

“chopsticks” method to catch the booster. To date, we have executed 11 Starship flight tests. We have also

scheduled a 12th flight test, which will debut the next generation Starship vehicle and Super Heavy booster. This

next-generation Starship introduces major changes for better orbital performance and reusability. We plan to

demonstrate key development milestones of catching the upper stage and demonstrating in-orbit propellant transfer

capabilities. These milestones will be the key unlocks for a rapidly reusable rocket that we expect will take hundreds

of thousands of tons of mass to orbit to drive growth in our Connectivity and AI segments, and allow us to develop

the lunar economy and eventually to reach Mars. We expect Starship to commence payload delivery to orbit in the

second half of 2026 following additional flight tests. For additional information about this risk, please refer to “Risk

Factors—Risks Related to Our Business—Any failure or delay in the development of Starship at scale or in

achieving the required launch cadence, reusability and capabilities thereafter would delay or limit our ability to

execute our growth strategy, including the deployment of next-generation satellites, global satellite-to-mobile

connectivity, and orbital AI compute, which could materially adversely affect our business, financial condition,

results of operations, and future prospects” in this prospectus.

Launch Costs and Cadence.Our launch costs and cadence underpin the foundational competitive advantage that

enables the performance of each of our segments. The reusability of our launch vehicles meaningfully reduces the

cost per kilogram to orbit by eliminating or limiting the need to manufacture new vehicles for every mission.

Reusability also enables higher launch cadence by shortening the time between flights, as vehicles can be rapidly

reflown after their return. These factors enable performance in our Connectivity segment by supporting faster and

more cost‑effective deployment of our satellite constellations. We expect they will support our AI segment as we

aim to deploy a large fleet of orbital AI compute. We expect continued enhancements to our launch infrastructure

and launch vehicles, including Starship, to drive cost down and throughput up, extending these benefits to our

businesses, as well as to our third‑party customers who rely on our launch capabilities. As we continue to reduce

launch costs and increase launch cadence, we expect to transform the rocket launch industry into airline-like

operations, enabling continuous and affordable access to space. Period-to-period comparisons of launch costs and

cadence are impacted by factors out of our control, including timing of delivery of customer payloads which impacts

94

the mix of customer and internal payloads and related financial reporting, or weather which can delay a launch from

one period to another.

Increasing Satellite Capacity. The scale, reliability, and capacity of our LEO broadband and mobile satellite

constellations drive our Connectivity segment’s growth and operating performance. In 2025, launching and

operating higher-throughput satellites supported Starlink’s service quality and customer reach by increasing

available network capacity and improving service consistency during peak usage periods. As of March 31, 2026, we

operated over 9,600 Starlink broadband and mobile satellites in Low-Earth Orbit, with the majority composed of our

second-generation, V2 Mini satellites. We expect to commence deploying our next-generation V3 satellites,

designed to offer one Tbps of downlink capacity per satellite, using Starship in the second half of 2026 and expect

that a single Starship launch will be capable of deploying up to 60 V3 satellites to LEO, representing a twenty-fold

increase in Starlink downlink capacity deployed relative to a Falcon 9 launch.

We also provide satellite-to-mobile connectivity, supplementing terrestrial networks and substantially reducing

mobile “dead zones” in approximately 30 countries. Since January 2025, we have grown our constellation from

approximately 360 mobile V1 Mobile satellites to approximately 650 mobile V1 Mobile satellites. Through this

constellation and in partnership with more than 30 mobile network operators, we provided data, over-the-top voice,

and messaging services to approximately 7.4 million monthly unique devices across approximately 30 countries.

During 2025, we also entered into agreements to acquire 65 MHz of spectrum in the United States as well as certain

global Mobile Satellite Service spectrum licenses from EchoStar for $19.6 billion of equity and cash consideration,

as described below under “—Liquidity and Capital Resources—Material Cash Commitments.” We expect the

spectrum acquisition to close in November 2027, subject to required regulatory approvals and other closing

conditions. We expect the wider bandwidth operations enabled by this spectrum purchase, together with our

authorization to deploy 7,500 satellites including with the 2GHz spectrum band, will provide stronger support for

current performance and potential future services, including broadband data and IoT connectivity, and is expected to

enable 5G connectivity.

These investments in satellite scale, per-satellite capacity, and expanded capabilities are instrumental to the growth

and operating performance of our Connectivity segment, enabling us to onboard new users while improving service

quality.

Increasing Starlink Brand Awareness and Acquiring New Subscribers. Our growth is driven in part by increased

global awareness of Starlink’s capabilities and our ability to convert that awareness into customer adoption. Trust,

visibility, and demonstrated reliability are central to customer acquisition, particularly for those in remote and

infrastructure-limited regions. Proven performance in rural, remote, and disaster-affected areas, along with strong

brand awareness, reinforces Starlink’s reputation as essential infrastructure, leading to higher adoption in new

markets.

As of March 31, 2026, we had over 9,600 Starlink broadband and mobile satellites in Low-Earth Orbit, operating the

world’s most advanced broadband constellation providing internet connectivity to approximately 10.3 million

Starlink Subscribers across 164 countries, territories, and other markets, collectively home to more than 3.3 billion

people. We are focused on growing the number of Starlink Subscribers by expanding our consumer distribution

network across thousands of authorized retail stores globally, and executing region-specific marketing campaigns to

increase brand awareness. By clearly demonstrating Starlink’s superior speed, low-latency, and ease of installation,

we expect to drive meaningful subscriber growth.

Increasing Enterprise Customer Adoption. As we continue to grow our Starlink constellation and bandwidth, we

see a large opportunity to grow the enterprise connectivity market by providing solutions that had not previously

been available. Our network is global and can provide primary connectivity for on-the-move applications as well as

a resilient backup option for enterprises serviced by land-based connectivity. We plan to deepen our penetration with

enterprise and government customers through direct, vertical-specific acquisition strategies. In recent years, we have

assembled dedicated sales and engineering teams to market and support fleet-wide conversions in aviation and

maritime, customized deployments for land mobility, which we expect to continue to grow as consumers who

experience Starlink begin to expect high-performance connectivity when traveling. We expect to enable more

customized deployments for land mobility across existing use cases such as commercial trucking fleets, and new

95

applications enabled by more connected devices. We also continue to develop specialized networks for secure

government applications via Starshield. By leveraging proven performance in mission-critical environments and

expanding through channel partners in select geographies, we expect to drive increased adoption among high-value

enterprise and government accounts.

Accelerating Investment in Growth and Innovation. We are simultaneously developing and scaling a wide range of

complex, capital‑intensive projects, including Starship and terrestrial and orbital AI compute. We believe speed is a

competitive advantage, and periodically we decide to increase and accelerate our investments. For example, in 2025

we accelerated our timeline for Starship development, increasing R&D in our Space segment to $3,004 million,

compared to $1,835 million in 2024. In our AI segment, in 2025 we successfully accelerated deployment of compute

for the development of Grok, increasing R&D in our AI segment to $5,064 million, compared to $1,176 million in

2024. We believe pursuing multiple ambitious programs in parallel enables us to compound advantages across our

vertically integrated innovation engine and unlock new large addressable markets over time. The timing of our

investments is not fixed and may accelerate based on technical progress, market opportunity, or resource

availability. As a result, our operating results, margins and profitability may fluctuate from period to period as we

continue to prioritize execution speed, capacity expansion, and technological leadership over near‑term margin

optimization. We believe that this approach maximizes long‑term value creation by allowing us to move faster than

competitors, scale earlier in emerging markets, and reinforce durable competitive advantages that we expect to

benefit our business over time.

Supply Chain and Manufacturing Efficiency for User Terminals. The operating performance of our Connectivity

segment depends in part on the cost and availability of user terminals at scale. We are vertically integrated across

terminal design, production, and support, including silicon, hardware, software, manufacturing, fulfillment, and

operations, which enables us to control our means of production as well as rapidly iterate to continuously improve

the performance of our user terminals and optimize product cost. Since our initial launch of our user terminal, we

have optimized the design of our phased-array antennas, our self-aligning antenna responsible for connecting user

equipment to our LEO satellite network, for manufacturability and high-volume scale. Over the past five years, we

have significantly lowered production costs and have scaled terminal output to approximately 200,000 terminals per

week. We plan to continue to further scale production significantly and make gains that improve margins, lower

customer barriers, and broaden addressable markets.

Scaling our AI Compute Rapidly and Efficiently.Our ability to rapidly and cost-effectively scale AI compute is a

significant driver of our competitiveness. We view scaling of compute capacity through a simple lens: power

availability and the powered shell together determine how quickly we can deploy compute, and our model and

serving stack in that powered shell determines how efficiently we convert that compute into useful tokens. In order

to scale our AI segment rapidly and efficiently, our strategy is extreme vertical integration, “from shovels to tokens.”

Power Availability and Powered Shells.We have demonstrated an industry-leading ability to rapidly deploy

large-scale data center infrastructure at unprecedented speed and cost efficiency. Our COLOSSUS and

COLOSSUS II data centers collectively provide approximately 1.0 gigawatt of compute power, with additional

power capacity available for data center operations. We brought the first cluster of COLOSSUS online in 122

days, repurposing the shell of an existing factory, and the first cluster of COLOSSUS II online even faster in 91

days. As an illustrative comparison, an industry benchmark to bring online a 100 megawatt greenfield data

center is approximately two years. We also demonstrated a significant improvement in cost efficiency,

achieving data center construction costs for COLOSSUS II that are considerably lower than industry

benchmarks on a per megawatt basis.

COLOSSUS and COLOSSUS II were brought online almost entirely through on-site power generation

capabilities that we designed, built, and deployed ourselves. We view our proven ability to construct power

infrastructure at this scale and speed as a significant competitive advantage. We partner closely with local

utilities to fund grid infrastructure expansions and access excess capacity, while proactively curtailing our grid

usage whenever required to prioritize community needs. Megapacks—utility-scale battery storage systems—

deliver critical redundancy and help stabilize operations during peak demand. Going forward, COLOSSUS II is

expected to be primarily powered by a dedicated natural gas power plant, supplemented over time by additional

grid capacity that we are directly funding through our local utility partners. Our comprehensive expertise across

the full infrastructure stack—from power procurement and on-site generation to distribution and advanced

96

cooling systems—enables us to translate available power into usable compute capacity with exceptional

efficiency. As we continue to scale and optimize, we expect to drive further improvements in Power Usage

Effectiveness. We expect these gains to accelerate the path from buildout to monetization.

AI Token Generation Efficiency. We are highly vertically integrated. We design, own or lease, and install all of

our powered shells and dedicated processor capacity. This full-stack ownership enables us to efficiently convert

power capacity into usable compute, precisely control cluster configuration, and operate a true end-to-end

system spanning infrastructure through to model deployment. Our operating performance depends on how

effectively we utilize deployed compute once capacity comes online—specifically, our ability to convert raw

infrastructure into reliable, high-throughput token generation at scale. Achieving this requires tight coordination

across model training and inference workflows, hardware configuration, and data center operations so that

utilization and throughput ramp efficiently as we expand. We believe we hold a meaningful efficiency

advantage by tightly integrating the model layer directly with the compute layer. Unlike third-party

environments that impose multiple abstraction layers, we run our serving stack close to the processors and

optimize serving, networking, and cluster configuration as a single unified system. This “model-to-compute”

integration reduces overhead, improves hardware utilization, and increases the proportion of available compute

that is converted into delivered output tokens. Output tokens represent the final generated response delivered to

the user, while total processing can be substantially higher when a request triggers additional inference-time

reasoning steps. Because we control workload scheduling and serving logic, we can prioritize high token

efficiency—intelligently balancing compute allocated to reasoning with strong final output to maintain or

improve response quality. This end-to-end control, combined with our sourcing relationships with leading

compute providers, gives us a performance-per-watt advantage and enables us to adopt new processor

generations at scale more rapidly through a repeatable playbook for reconfiguration and recommissioning.

Orbital AI Compute Has the Potential to Massively Increase Our Ability to Scale Our AI Compute,

Accelerate Our Pace, and to Be More Cost Effective Relative to Terrestrial Options. We believe we are the

only company with a commercially viable path to building orbital AI compute at scale. This is underpinned by

our unique ability to launch substantial mass into orbit cost efficiently through reusable rockets and manufacture

secure, reliable, and high performance satellites at low cost and high volume. We plan to develop orbital data

centers to enable scaling of compute capacity for us and our customers that is independent of terrestrial power

infrastructure constraints. Space offers the potential to access virtually limitless power and an operating

environment that supports sustained high‑density compute, including structural advantages for power

generation, cooling, and uninterrupted operations as capacity grows. We plan to employ a modular shell

approach built around our scalable satellite constellation, which enables compute capacity to be deployed and

expanded efficiently as capacity requirements grow. The architecture also supports shorter refresh cycles at the

token layer, as we can upgrade compute as successive chip generations arrive, increasing token output per unit

of installed capacity. Our goal over time is to launch 100 gigawatts of compute to space each year. If operated

continuously, the generation resources used to support 100 gigawatts of compute could generate approximately

one-fifth of the annual power production in the United States, which was 4.4 thousand terawatt hours in 2025,

according to the U.S. Energy Information Administration (EIA). We expect space‑based compute to massively

increase AI compute scale, while also improving token economics.

Ability to Increase Revenue from our Consumer User Base. Our performance depends in part on our ability to

effectively increase revenue from our over 1.3 billion accounts active in the last twelve months ended March 2026,

including approximately 550 million monthly active AI users across Grok and X through multiple complementary

monetization channels:

Growing our Advertising Platform.Advertising remains a core monetization channel for our AI segment, with

revenue driven by our ability to deliver highly relevant ads. We aim to grow advertising revenue per user by

strengthening performance advertising, expanding AI‑driven targeting and measurement, and introducing richer

ad formats and creative tools. A central focus of ours is making ads feel like content—contextually relevant,

aligned with user interests, and integrated into real‑time conversations. Grok increasingly supports this strategy

by helping advertisers with campaign creation, creative optimization, and alignment with trending topics and

user intent. While these factors help us drive advertising revenue, the pricing of our advertising products is also

affected by other factors, including the global economy and the highly competitive nature of our industry. We

97

believe continued investment in AI‑powered advertising will further improve advertiser ROI while further

enhancing user experience.

Conversion of Users to Paid Subscribers. In parallel, we are focused on converting a greater portion of our user

base into paying subscribers through our X subscription (Premium and Premium+) and Grok subscription

offerings. Subscribers benefit from enhanced functionality, exclusive features, and access to our latest AI

models. As of March 31, 2026, we reached approximately 6.3 million active paid subscribers, which was

comprised of approximately 4.4 million X Premium and Premium+ paid subscribers and approximately 1.9

million SuperGrok, SuperGrok Heavy and SuperGrok Lite paid subscribers. We plan to continue adding new

features and functionality while releasing increasingly capable Grok models to increase the penetration rate of

our subscriber base. Our AI segment has demonstrated exceptional model velocity: since launching Grok, we

have developed leading frontier models at a far faster rate of innovation than others. We believe this pace of

innovation strengthens the value proposition of our subscription offerings and supports long‑term subscriber

growth.

Progress Toward the Everything App and New Monetization Channels. We aim to evolve X into an

“Everything App,” integrating real-time information, communications, media, payments, banking, commerce

and more within one consumer experience. This can increase the usefulness of X, and therefore increase the

usage and monetization potential of X. We have rapid product launch velocity, with a frequent cadence of new

features and products launched since 2023, including features such as long‑form video, improved group

interactions, and creator tools. We plan to further broaden the value proposition of X through offerings like

Money, a product we launched in beta in November 2025, which aims to expand platform utility by enabling

payments and other financial services. We released X Chat in November 2025, which features end-to-end

encryption and has no connection to advertising, unlike other services. We intend to further embed Grok

throughout the platform to enhance discovery, analysis of posts, user support, and personalization, making core

workflows more useful and reducing friction for users to adopt paid features.

Growing Enterprise and Government Adoption of Our AI Offerings. Our future growth and financial performance

depend in part on our ability to increase adoption and usage of our AI offerings among enterprise and government

customers. We have launched Grok Business, Grok Enterprise, Grok API, and xAI Gov, products that we believe

will be attractive to enterprises and governments, and we expect substantial opportunities to acquire new customers.

We are also partnering with Cursor to advance Grok and potentially to create jointly-owned coding and knowledge

work AI models, trained on our compute infrastructure. Over time, we also believe enterprises and governments will

present significant opportunities for revenue expansion as they deploy our models more broadly across their

organizations, adopt new capabilities, and build and operate solutions using our API. We also intend to continue to

offer our compute infrastructure to third-party customers. Our ability to realize these expansion opportunities

depends on continued innovation, reliable performance, and meeting evolving technical, security, and compliance

requirements.

Components of Results of Operations

Description of Our Segments

Space

Revenue - Space

Space segment generates revenue primarily through (i) Launch Services for the deployment of payloads to their

intended orbits for both commercial and government customers utilizing Falcon 9 and Falcon Heavy, and (ii)

Launch and Development for the development of spacecraft and provision of launch and mission services for

government agency space programs utilizing Falcon 9, Falcon Heavy, Starship, and Dragon. Launch Services

revenue is derived from fixed-price contracts that range from one to five years. Launch and Development revenue is

derived from fixed-price contracts that can range from one to fourteen years.

98

The Company recognizes Launch Services revenue at a point in time, due to the interchangeability of flight

hardware and minimal unique engineering costs. Revenue and costs are deferred and not recognized until upon the

launch or deployment of the customer’s payload to their intended orbit.

The Company recognizes Launch and Development revenue over time as the Company’s performance on the

contract creates an asset with no alternative use and the Company has an enforceable right to payment for

performance to date. The Company measures progress on these contracts using the cost-to-cost input method, which

the Company believes represents the most appropriate measure towards satisfaction of its performance obligation.

For launches of our Starlink satellites, the Company does not recognize any inter-segment revenue, rather those

launch costs are capitalized in satellites in Property, plant, and equipment, net.  We allocate a significant amount of

launch capacity to our Connectivity segment, and expect to allocate a significant amount to our AI segment in the

future. Our Space segment revenue only reflects our customer launches and customer activities.

Revenue from Launch Services recognized at point in time and revenue from Launch and Development recognized

over time as a percentage of total Space segment revenue are as follows:

	Three Months Ended March 31,		Year Ended December 31,
	2026	2025	2025	2024	2023
Launch Services ...............................	53.3%	65.4%	63.0%	68.2%	55.2%
Launch & Development ...................	46.7%	34.6%	37.0%	31.8%	44.8%
Space................................................	100.0%	100.0%	100.0%	100.0%	100.0%

We expect Space revenue growth to continue to be lower than total company revenue growth as our internal

business continues to absorb most of the growth in our launch capacity. In addition, we expect Launch and

Development to represent a larger portion of our Space revenue as we continue to serve our long-term contracts for

our government customers. From period to period, Space revenue will vary based on the mix of launches used for

customers and our own businesses.

Expenses - Space

Cost of Revenue

The Company’s Falcon 9 and Falcon Heavy are composed of boosters (also known as first stages), second stages,

Merlin engines, and fairings. Boosters, fairings, and Merlin engines are reusable and are classified as property, plant,

and equipment and are depreciated to cost of revenue. The second stages are not reusable and are recorded to cost of

revenue when they are launched for Launch Services revenue transactions or assigned for Launch and Development

revenue transactions. Dragon is comprised of a fully reusable capsule that is classified asProperty, plant, and

equipment, net and is depreciated to cost of revenue. Starship is comprised of a booster, ship, and Raptor engines

and is currently in the development stage. A majority of Starship costs are currently expensed to Research and

development as incurred. Raptor engines are expensed when used in test flights.   

Space segment’s cost of revenue includes second stages flown related to the Company’s Falcon 9 and Falcon Heavy

launches, launch operations and overhead, depreciation (inclusive of booster, Merlin engine, and fairing

depreciation), employee compensation costs (including salaries, benefits, and share-based compensation) for our

operations teams, launch testing and overhead, engineering costs, inventory excess and obsolescence, shared costs

incurred in the production of launch hardware, and ongoing product support. 

We expect Space cost of revenue to increase both in absolute dollars and as a percentage of revenue based on our

expected mix of Launch Services and Launch and Development. From period to period, Space segment cost of

revenue will vary based on the mix of customer and internal launches.

Research and Development

Space segment’s research and development (“R&D”) expenses mainly relate to the development, build, and testing

of Starship. Starship costs consist of test flight hardware, Raptor engines, employee compensation costs (including

99

salaries, benefits, and share-based compensation), tooling and equipment expenses, depreciation for R&D

equipment, and allocated overhead. R&D also includes certain expenses related to the development of features and

modules created through engineering services for the Company’s Falcon vehicles, where the Company retains the

associated intellectual property.

We expect Space research and development to increase both in absolute dollars and as a percentage of revenue in

2026, as we invest in the development and commercialization of Starship, and to moderate both in absolute dollars

and as a percentage of revenue once Starship is commercialized by delivering payload to orbit. At

commercialization, Starship costs generally will be capitalized and then depreciated in cost of revenue of the

segment associated with the payload delivered.

Selling, General, and Administrative

Space segment’s selling, general, and administrative (“SG&A”) expenses include allocated employee compensation

costs (including salaries, benefits, and share-based compensation) for our sales, facilities, legal, finance, information

technology, human resources, and other administrative employees, depreciation, and corporate aircraft costs.

We expect Space segment's SG&A to increase in absolute dollars to support growth of our business, and to decrease

as a percentage of revenue as we continue to work to reduce operating costs as a percentage of revenue.

Impairment

Space impairment includes impairment losses on fixed assets due to anomalies on the Company’s flight vehicles and

launch sites, which occur outside our normal business operations.

Connectivity

Revenue - Connectivity

Connectivity segment generates revenue from (i) the broadband and mobile connectivity services provided through

Starlink and (ii) the sale of the Starlink Kit (inclusive of the terminal). The Company provides connectivity services

and Starlink Kits to consumers or enterprise and government customers.

The Company recognizes revenue from broadband and mobile connectivity services over time as the customer

simultaneously receives and consumes the benefits provided. The Company generates service revenue from (i)

fixed-price services that require advance or recurring monthly payments by the customer or (ii) variable-priced

services based on actual data consumption. The amounts received from customers for advanced payments for

broadband and mobile connectivity services are recognized either ratably over the subscription term or based on

actual data consumption. The Company’s broadband contracts are generally month-to-month and the revenue

recognized for these recurring consumer customers is equal to the amount billed in that month.  The Company’s

mobile connectivity agreements are generally multi-year contractual obligations that range from one to five years,

although the customer can generally terminate at any time.

The Company recognizes revenue over time for certain contracts related to our Starshield business that are multi-

year in nature. For revenue that is recognized over time, we use the cost-to-cost input method. The Company records

revenue based upon costs (such as materials and labor hours) incurred to date relative to the total estimated cost at

completion.

The Company records revenue for the Starlink Kit upon delivery to the customer, or in the instance of certain

enterprise customers, when it is installed. Starlink Kit revenue is reported net of sales returns, credits, and

chargebacks.

100

Expenses - Connectivity

Cost of Revenue

Connectivity segment’s cost of revenue includes depreciation (inclusive of launch, satellite, and ground

infrastructure costs), Starlink Kit costs, shipping and handling costs, ground operating expenses, employee

compensation costs (including salaries, benefits, and share-based compensation) for our engineering and operations

teams, payment processor fees, warranty expense, inventory excess and obsolescence, and customs and duties.

We expect Connectivity cost of revenue to increase in absolute dollars as we grow our revenue, and to decrease as a

percentage of revenue as we continue to drive efficiencies in our next-generation satellites, Starlink Kits, and ground

infrastructure.

Research and Development

Connectivity segment’s R&D expenses mainly relate to the development, build, and testing of our next-generation

satellites, Starlink Kits, and ground infrastructure. These costs include employee compensation costs (including

salaries, benefits, and share-based compensation), contractor compensation expenses, equipment lease expenses,

depreciation for R&D equipment, and allocated overhead.

We expect Connectivity research and development to increase in absolute dollars as we grow our revenue, and to

decrease as a percentage of revenue as we scale our business.

Selling, General, and Administrative

Connectivity segment’s SG&A expenses include allocated employee compensation costs (including salaries,

benefits, and share-based compensation) for our sales, facilities, legal, finance, information technology, human

resources, and other administrative employees, licensing and regulatory fees, marketing expenses, depreciation, and

bad debt expense.

We expect Connectivity SG&A to increase in absolute dollars and as a percentage of revenue in 2026 as we

introduce marketing spend to support growth of our business, and to decrease as a percentage of revenue over time

as we continue to work to reduce operating costs as a percentage of revenue.

Impairment

Connectivity impairment includes costs related to discontinuation of a product line for Starlink Kits that is non-

recurring.

AI

Revenue - AI

AI segment generates revenue from the sale of digital platform services, including advertising, subscription, and

licensing services offered to consumers and enterprise customers. 

The Company generates revenue from (i) the sale of ad products displayed on its X platform, and (ii) providing AI

solutions and infrastructure, which includes subscription-related offerings, data licensing arrangements, and API

access to Grok models.

Revenue for advertising services is recognized in the period when advertising is delivered as evidenced by a person

engaging with an ad on the Company’s platforms in a manner satisfying the types of engagement selected by the

advertisers.  The Company’s contract terms for advertising services are typically cancellable short-term

arrangements. We experience seasonality in our advertising revenues. Overall advertising spend tends to be highest

in the fourth quarter of each year due in large part to end-of-year advertiser spending and lowest in the first quarter

of each year. 

101

Revenue for AI solutions and infrastructure includes: (i) premium subscriptions on X and Grok which is recognized

ratably over the period of the subscription term (ranging from month-to-month to one year), (ii) data licensing

revenue which is generally recognized ratably over the period (from month-to-month to two years) in which the

Company provides data as the customer consumes and benefits from the use of the licensed data, (iii) revenue from

providing API access to Grok models recognized ratably over the contract term (typically month-to-month or up to

one year) for stand-ready access or as services are consumed for usage based arrangements.

Expenses - AI

Cost of Revenue

AI segment’s cost of revenue includes infrastructure costs, revenue share expenses, payment processor fees,

payments to creators, amortization of acquired intangible assets, and allocated labor and overhead costs.

Infrastructure costs consist primarily of costs related to data center facilities, including lease and hosting costs,

related support, maintenance, energy, and bandwidth costs, depreciation of servers and networking equipment,

public cloud hosting costs, and employee compensation costs (including salaries, benefits, and share-based

compensation) for our operations teams.

We expect AI cost of revenue to increase in absolute dollars as we grow our revenue, and to decrease as a

percentage of revenue as we monetize our products and as we expand our service offerings for AI solutions.

Research and Development

AI segment’s R&D expenses mainly relate to the training of Grok, our leading frontier model, development, build,

and testing of our next-generation AI-enabled products and data center costs to train AI-enabled products. These

costs include cloud computing expenses, employee compensation expenses (including salaries, benefits, and share-

based compensation), power generation costs, and depreciation of data center assets, including processors,

equipment lease expenses, and networking equipment.

We expect AI R&D expenses to increase, both in absolute dollars and as a percentage of revenue, as we invest in

compute infrastructure for Grok. Additionally, AI R&D expenses may increase as a result of the compute agreement

with Cursor.

Selling, General, and Administrative

AI segment’s SG&A expenses consist primarily of employee compensation expenses (including salaries, benefits,

and share-based compensation) for our sales, sales support, marketing, finance, legal, information technology,

human resources and other administrative employees. In addition, SG&A expenses include fees and costs for

professional services, including consulting, content moderation, third-party legal and accounting services and

facilities costs and other supporting overhead costs that are not allocated to other departments.

We expect AI SG&A to increase in absolute dollars to support growth of our business, and to decrease as a

percentage of revenue as we continue to work to reduce operating costs as a percentage of revenue. Additionally, AI

SG&A may increase as a result of the compute agreement with Cursor.

Restructuring Charges

AI restructuring charges are the result of the acquisition of Twitter in October 2022 by X Holdings. The charges

include workforce restructuring for former Twitter employees, as well as impairment and early termination penalties

as a result of consolidation of Twitter’s various office leases.

Impairment

AI impairment includes a one-time impairment of the Twitter brand when Twitter was rebranded to X in July 2023. 

102

Other Corporate Expenses

Interest Expense

Interest expense includes interest expense related to our borrowings, amortization of associated debt issuance costs,

undrawn fees, and finance leases. Interest expense is reflected net of capitalized interest.

Interest Income

Interest income includes interest income earned on cash and cash equivalents and marketable securities, and

dividend income from our investments in mutual funds.

Other Income (Expense), Net

Other income (expense), net consists of gain or loss on digital assets, gain or loss on foreign currency transactions,

and loss on extinguishment of debt.

Provision for (Benefit from) Income Taxes

The provision for (benefit from) income taxes consists primarily of income taxes in certain federal, state, local and

foreign jurisdictions in which we conduct business. Foreign jurisdictions typically have different statutory tax rates

from those in the United States. Accordingly, our effective tax rates may vary depending on the impact of the

valuation allowance as well as the relative proportion of foreign income to domestic income, generation of tax

credits, changes in the valuation of our deferred tax assets and liabilities, and changes in tax laws.

Comparison of the three months ended March 31, 2026 and 2025

Consolidated Results of Operations

The following table sets forth our consolidated financial statements data for the periods indicated:

	Three Months Ended March 31,		2026 vs. 2025 Change
(in millions)	2026	2025	$ Change	% Change
Revenue...............................................................	$4,694	$4,067	$627	15.4%
Costs and expenses
Cost of revenue..............................................	2,388	1,962	426	21.7%
Research and development.............................	3,514	1,557	1,957	125.7%
Selling, general, and administrative...............	746	493	253	51.3%
Restructuring charges (credits).......................	(11)	4	(15)	NM
Impairment.....................................................	—	24	(24)	NM
Total costs and expenses...........................	6,637	4,040	2,597	64.3%
Income (loss) from operations............................	(1,943)	27	(1,970)	NM
Interest expense...................................................	(664)	(447)	(217)	48.5%
Interest income....................................................	213	117	96	82.1%
Other expense, net...............................................	(1,876)	(211)	(1,665)	789.1%
Loss before income taxes....................................	(4,270)	(514)	(3,756)	730.7%
Provision for income taxes..................................	6	14	(8)	(57.1)%
Net loss................................................................	$(4,276)	$(528)	$(3,748)	709.8%

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

103

Revenue

Revenue for the three months endedMarch 31, 2026increased by $627 million, or 15.4%, compared to thethree

months endedMarch 31, 2025. This increase was primarily due to an increase in revenue from our Connectivity

segment of $782 million as our Starlink Subscriber base continued to grow as well as an increase in revenue from

our AI segment of $91 million from higher X and Grok subscriptions, partially offset by a decrease in revenue from

our Space segment of $246 million due to lower Launch Services missions and timing of work for government

contracts.

Cost of Revenue

Cost of revenue for the three months endedMarch 31, 2026increased by $426 million, or 21.7%, compared to the

prior three months endedMarch 31, 2025. This increase was primarily due to an increase in costs in our

Connectivity segment of $437 million driven by an increase in depreciation related to the number of satellites placed

into orbit and higher operating costs of $5 million in our AI segment, partially offset by a decrease in cost of revenue

from our Space segment of $16 million due to less customer launches.

Research and Development

Research and development expense for the three months endedMarch 31, 2026increased by $1,957 million, or

125.7%, compared to the prior three months endedMarch 31, 2025. This increase was primarily due to higher costs

in our AI segment of $1,471 million driven by depreciation of GPU hardware, and the cost of cloud computing and

data center infrastructure expenses as a result of our AI data center expansions and higher costs from our Space

segment of $404 million driven by accelerated investment in our Starship vehicle and related facilities.

Selling, General, and Administrative

Selling, general, and administrative expense for the three months endedMarch 31, 2026increased by $253 million,

or 51.3%, compared to the prior three months endedMarch 31, 2025. This increase was primarily due to higher

employee-related costs and professional fees for our AI segment of $163 million as our AI business grew rapidly,

higher marketing and international expansion costs of $79 million and $23 million, respectively, for our

Connectivity segment. These increases were partially offset by lower expenses of $18 million in our Space segment.

Restructuring Charges (Credits)

Restructuring charges (credits) for the three months endedMarch 31, 2026decreased by $15 million compared to

the prior three months endedMarch 31, 2025.  This decrease was primarily due to change in estimated settlement

amounts for former Twitter employees as part of the workforce reduction program implemented in 2022.

Impairment

Impairment for the three months endedMarch 31, 2026decreased by $24 millioncompared to the prior three

months endedMarch 31, 2025. The impairment in the three months endedMarch 31, 2025 was related to a post-

landing anomaly in our Space segment.  There was no impairment for the three months ended March 31, 2026.

Income (Loss) from Operations

Income (loss) from operations for the three months endedMarch 31, 2026decreased by $1,970 million compared to

the prior three months endedMarch 31, 2025 driven by the factors described above.

Interest Expense

Interest expense for the three months endedMarch 31, 2026increased by $217 million, or 48.5%, compared to the

prior three months endedMarch 31, 2025. This increase was primarily due to additional debt raised by the Company

and other financing arrangements entered into during the period by our AI segment.

104

Interest Income

Interest income for the three months endedMarch 31, 2026increased by $96 million, or 82.1%, compared to the

prior three months endedMarch 31, 2025. This increase was primarily due to an increase in interest income earned

from cash equivalents and marketable securities.

Other Income (Expense), Net

Other expense, net for the three months endedMarch 31, 2026increased by $1,665 million, compared to the prior

three months endedMarch 31, 2025. This increase was primarily due to the loss on extinguishment of debt and

unrealized loss on digital assets.

Provision for (Benefit from) Income Taxes

Provision for income taxes for the three months endedMarch 31, 2026decreased by $8 million compared to the

prior three months endedMarch 31, 2025. This decrease was primarily due to the change in the mix of our

jurisdictional earnings subject to different tax rates.

Net Income (Loss)

Net loss for the three months endedMarch 31, 2026increased by $3,748 million compared to the prior three months

endedMarch 31, 2025 driven by the factors described above.

Segment Results

Space

	Three Months Ended March 31,		2026 vs. 2025 Change
(in millions)	2026	2025	$ Change	% Change
Revenue.............................................................	$619	$865	$(246)	(28.4)%
Costs and expenses
Cost of revenue...................................................	281	297	(16)	(5.4)%
Research and development.............................	930	526	404	76.8%
Selling, general, and administrative...............	70	88	(18)	(20.5)%
Impairment.....................................................	—	24	(24)	NM
Total costs and expenses................................	$1,281	$935	$346	37.0%
Loss from operations.................................	$(662)	$(70)	$(592)	845.7%

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the three months endedMarch 31, 2026decreased$246 million, or 28.4%, compared to the prior three

months endedMarch 31, 2025. This decrease was primarily driven by a decrease in Launch Services revenue of

$236 million and a decrease of $10 million in Launch and Development revenue. The decrease in Launch Services

revenue is due to a decrease in customer launches period over period. While total Falcon launches increased by 4

from 36 for the three months ended March 31, 2025 to 40 for the three months ended March 31, 2026, Launch

Services missions decreased by 4 over the same period. Launch and Development revenue decreased due to timing

of work performed on government contracts. 

Cost of Revenue

Cost of revenue for the three months endedMarch 31, 2026decreased by $16 million, or 5.4%, compared to the

prior three months endedMarch 31, 2025. This decrease was primarily due to the decrease in customer launches and

timing of work on government contracts of $34 million, offset by an increase of $10 million in inventory excess and

obsolescence reserves and $10 million in launch hardware disposals for damaged Falcon fairings.

105

Research and Development

Research and development for the three months endedMarch 31, 2026increased by $404 million, or 76.8%,

compared to the prior three months endedMarch 31, 2025. This increase was primarily driven by higher production

costs of $194 million, higher engineering costs of $95 million, and higher test and launch costs of $62 million, due

to the accelerated investment in development of the Starship vehicle and continued development of production and

launch facilities to support future Starship launches.

Selling, General, and Administrative

Selling, general, and administrative for the three months endedMarch 31, 2026decreased by $18 million, or 20.5%,

compared to the prior three months endedMarch 31, 2025. This decrease was primarily due to lower allocated

general and administrative overhead of $13 million.

Impairment

Impairment for the three months endedMarch 31, 2026decreased by $24 million compared to the prior three

months endedMarch 31, 2025. This decrease was primarily due to a non-recurring impairment loss on a Falcon 9

booster due to a post-landing anomaly during the three months endedMarch 31, 2025. There was no impairment for

the three months ended March 31, 2026.

Loss from Operations

Space loss from operations for the three months endedMarch 31, 2026increased by $592 million compared to the

prior three months endedMarch 31, 2025 driven by the factors described above.

Connectivity

	Three Months Ended March 31,		2026 vs. 2025 Change
(in millions)	2026	2025	$ Change	% Change
Revenue.............................................................	$3,257	$2,475	$782	31.6%
Costs and expenses
Cost of revenue..............................................	$1,651	$1,214	$437	36.0%
Research and development.............................	205	123	82	66.7%
Selling, general, and administrative...............	213	105	108	102.9%
Total costs and expenses................................	$2,069	$1,442	$627	43.5%
Income from operations.............................	$1,188	$1,033	$155	15.0%

Revenue

Revenue for the three months endedMarch 31, 2026increased by $782 million, or 31.6%, compared to the prior

three months endedMarch 31, 2025. This increase was primarily driven by an increase of $656 million in revenue

from our consumer subscribers, composed of 104.7% growth in Starlink Subscribers, offset by an 22.9%decline in

Starlink Subscriber ARPU, primarily due to international expansion and the addition of lower priced service plans.

In addition, enterprise and government revenue had an increase of $126 million primarily driven by the growth in

our aviation, maritime, and other enterprise business of $209 million, our mobile connectivity business of $85

million, partially offset by a decrease of $175 million in our government connectivity business.

Cost of Revenue

Cost of revenue for the three months endedMarch 31, 2026increased by $437 million, or 36.0%, compared to the

prior three months endedMarch 31, 2025. This increase was primarily due to higher depreciation of $276 million

from capitalized launch and satellite costs, higher operating expenses of $140 million mainly driven by ground

operating costs of $50 million, customer support and installation costs of $42 million, payment processor fees of $19

million, freight costs of $15 million, and warranty costs of $12 million.

106

Research and Development

Research and development for the three months endedMarch 31, 2026increased by $82 million, or 66.7%,

compared to the prior three months endedMarch 31, 2025. This increase was primarily due to higher costs for the

next-generation production development of satellites of $62 million, Starlink Kits of $8 million, and ground

equipment of $14 million.

Selling, General, and Administrative

Selling, general, and administrative for the three months endedMarch 31, 2026increased by $108 million, or

102.9%, compared to the prior three months endedMarch 31, 2025. This increase was primarily driven by higher

marketing costs of $79 million and higher international expansion costs of $23 million, partially offset by lower bad

debt expense of $9 million.

Income from Operations

Connectivity income from operations for the three months endedMarch 31, 2026increased by $155 million, or

15.0%, compared to the prior three months endedMarch 31, 2025 driven by the factors described above.

AI

	Three Months Ended March 31,		2026 vs. 2025 Change
(in millions)	2026	2025	$ Change	% Change
Revenue.............................................................	$818	$727	$91	12.5%
Costs and expenses
Cost of revenue..............................................	456	451	5	1.1%
Research and development.............................	2,379	908	1,471	162.0%
Selling, general, and administrative...............	463	300	163	54.3%
Restructuring charges.....................................	(11)	4	(15)	NM
Total costs and expenses...........................	$3,287	$1,663	$1,624	97.7%
Loss from operations.................................	$(2,469)	$(936)	$(1,533)	163.8%

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the three months endedMarch 31, 2026increased by $91 million, or 12.5%, compared to the prior three

months endedMarch 31, 2025 due to the increase in AI solutions and infrastructure revenue of $191 million, offset

by decrease in advertising revenue of $100 million. The increase in AI solutions and infrastructure was primarily

due to an increase in Grok and X subscription revenue of $177 million and an increase in data licensing

arrangements of $12 million. The decrease in advertising revenue is due to an overhaul of the Company’s

advertising platform which impacted ad sales for a short period of time during the rebuild.

Cost of Revenue

Cost of revenue for the three months endedMarch 31, 2026increased by $5 million, or 1.1%, compared to the prior

three months endedMarch 31, 2025. This increase was primarily due to an increase in revenue share and content

creator expenses of $71 million, and higher payment processing fees of $18 million, partially offset by a decrease in

amortization expenses of technology intangibles of $89 million that were fully amortized during 2025.

Research and Development

Research and development for the three months endedMarch 31, 2026increased by $1,471 million, or 162.0%,

compared to the prior three months endedMarch 31, 2025. This increase was primarily due to higher GPU

depreciation expense of $908 million, and higher cloud computing anddata center infrastructure expenses of $301

107

million associated with the continued build out of our compute infrastructure, as well as higher employee

compensation expenses (including salaries, benefits, and share-based compensation) of $262 million.

Selling, General, and Administrative

Selling, general, and administrative for the three months endedMarch 31, 2026increased by $163 million, or

54.3%, compared to the prior three months endedMarch 31, 2025. This increase was primarily due to higher

employee compensation expenses (including salaries, benefits, and share-based compensation) of $148 million as

we continue to expand our AI business and higher legal expenses of $33 million, partially offset by a decrease in

facilities and general and administrative costs of $18 million.

Restructuring Charges (Credits)

Restructuring charges (credits) for the three months endedMarch 31, 2026decreased by $15 million compared to

the prior three months endedMarch 31, 2025. This decrease was primarily due to a change in estimated settlement

amounts for former Twitter employees as part of the workforce reduction program implemented in 2022.

Loss from Operations

AI loss from operations for the three months endedMarch 31, 2026increased by$1,533 million, or 163.8%,

compared to the prior three months endedMarch 31, 2025 driven by the factors described above.

Comparison of the Years Ended December 31, 2025 and 2024

Consolidated Results of Operations

The following table sets forth our consolidated statements of operations data for the periods indicated:

	Year Ended December 31,		2025 vs. 2024 Change
(in millions)	2025	2024	$ Change	% Change
Revenue...............................................................	$18,674	$14,015	$4,659	33.2%
Costs and expenses
Cost of revenue..............................................	9,451	7,996	1,455	18.2%
Research and development.............................	8,643	3,464	5,179	149.5%
Selling, general, and administrative...............	2,644	1,813	831	45.8%
Restructuring charges.....................................	487	213	274	128.6%
Impairment.....................................................	38	63	(25)	(39.7)%
Total costs and expenses...........................	21,263	13,549	7,714	56.9%
Income (loss) from operations............................	(2,589)	466	(3,055)	NM
Interest expense...................................................	(1,945)	(1,580)	(365)	23.1%
Interest income....................................................	492	371	121	32.6%
Other income, net................................................	(177)	985	(1,162)	NM
Income (loss) before income taxes......................	(4,219)	242	(4,461)	NM
Provision for (benefit from) income taxes..........	718	(549)	1,267	NM
Net income (loss)................................................	$(4,937)	$791	$(5,728)	NM

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the year ended December 31, 2025increased by $4,659 million, or 33.2%, compared to the prior year

ended December 31, 2024. This increase was primarily due to an increase in revenue from our Connectivity segment

of $3,788 millionas our Starlink Subscriber base continued to grow as well as our Connectivity enterprise and

government sales, and increases in revenue from our Space segment of $290 million due to increases in Launch and

108

Development revenue for work performed on government contracts, and an increase in revenue from our AI

segment of $581 million as advertising, Grok and X subscriptions, and data licensing arrangements grew.

Cost of Revenue

Cost of revenue for the year ended December 31, 2025increased by $1,455 million, or 18.2%, compared to the prior

year ended December 31, 2024. This increase was primarily due to an increase in costs in our Connectivity segment

of $1,153 million driven by higher depreciation as the number of satellites placed into orbit grew and higher

operating expenses, and higher infrastructure and cloud computing costs of $491 million in our AI segment, partially

offset by a decrease in cost of revenue from our Space segment of $189 million due to the increased reusability of

our Falcon launch vehicles resulting in lower depreciation. 

Research and Development

Research and development expense for the year ended December 31, 2025increased by $5,179 million, or 149.5%,

compared to the prior year ended December 31, 2024. This increase was primarily due to higher R&D costs in our

AI segment of $3,888 million driven by the depreciation of GPU hardware and the cost of cloud computing as a

result of our AI data center expansions and higher R&D costs from our Space segment of $1,169 milliondriven by

accelerated investment in our Starship vehicle.

Selling, General, and Administrative

Selling, general, and administrative expense for the year ended December 31, 2025increased by $831 million, or

45.8%, compared to the prior year ended December 31, 2024. This increase was primarily due to higher employee

and facilities-related costs and higher legal expenses for our AI segment of $722 million as our AI business grew

rapidly, and higher marketing and international expansion costs of $53 million and $37 million, respectively, for our

Connectivity segment. These increases were partially offset by lower allocated general and administrative overhead

in our Space segment.

Restructuring Charges

Restructuring charges for the year ended December 31, 2025increased by $274 million, or 128.6%, compared to the

prior year ended December 31, 2024.  This increase was primarily due to additional expense related to the settlement

to former Twitter employees as part of the workforce reduction program implemented in 2022.

Impairment

Impairment for the year ended December 31, 2025decreased by $25 million, or 39.7%, compared to the prior year

ended December 31, 2024. The decrease was primarily related to a discontinuation of a Starlink Kit production line

in our Connectivity segment that occurred during the year ended December 31, 2024 with no impairment in 2025,

partially offset by an increase in impairment in our Space Segment during the year ended December 31, 2025

primarily related to a post-landing anomaly.

Income (Loss) from Operations

Income (loss) from operations for the year ended December 31, 2025decreased by $3,055 million compared to the

prior year ended December 31, 2024 driven by the factors described above.

Interest Expense

Interest expense for the year ended December 31, 2025increased by $365 million, or 23.1%, compared to the prior

year ended December 31, 2024. This increase was primarily due to new term loans and senior notes entered into by

the Company and other financing arrangements for GPUs entered into during the year by our AI segment.

109

Interest Income

Interest income for the year ended December 31, 2025increased by $121 million, or 32.6%, compared to the prior

year ended December 31, 2024. This increase was primarily due to an increase in dividend income earned from

marketable securities and cash equivalents.

Other Income (Expense), Net

Other income (expense), net for the year ended December 31, 2025decreased by $1,162 million, compared to the

prior year ended December 31, 2024. This decrease was primarily due to an unrealized loss on digital assets.

Provision for (Benefit from) Income Taxes

Provision for income taxes for the year ended December 31, 2025 increased by $1,267 million compared to the prior

year ended December 31, 2024. The increase was primarily due to a partial valuation allowance release in 2024 and

the establishment of a valuation allowance in 2025. For the year ended December 31, 2024, the Company released a

partial valuation allowance on the Company’s U.S. deferred tax assets. As of December 31, 2024, the Company

forecasted $659 million of deferred tax assets related to U.S. R&D credits would be utilized in the future. For the

year ended December 31, 2025, as a result of the enactment of the One Big Beautiful Bill Act (Public Law No.

119-21), we assessed the realizability of our deferred tax assets and reversed the benefit that was recognized for the

year ended December 31, 2024.

Net Income (Loss)

Net income (loss) for the year ended December 31, 2025decreased by $5,728 million compared to the prior year

ended December 31, 2024 driven by the factors described above.

Segment Results

Space

	Year Ended December 31,		2025 vs. 2024 Change
(in millions)	2025	2024	$ Change	% Change
Revenue...............................................................	$4,086	$3,796	$290	7.6%
Costs and expenses
Cost of revenue..............................................	1,352	1,541	(189)	(12.2)%
Research and development.............................	3,004	1,835	1,169	63.7%
Selling, general, and administrative...............	349	375	(26)	(6.9)%
Impairment.....................................................	38	24	14	61.5%
Total costs and expenses...........................	$4,743	$3,775	$968	25.7%
Income (loss) from operations............................	$(657)	$21	$(678)	NM

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the year ended December 31, 2025increased by $290 million,or 7.6%, compared to the prior year

ended December 31, 2024.  Launch Services revenue remained relatively flat year over year, while Launch and

Development revenue increased by $298 million.  The Launch and Development revenue increase was primarily

driven by increased revenue for an extended contract with NASA for additional Cargo Resupply Services (CRS)

missions to the International Space Station and increased revenue from a U.S. Department of War contract. While

total Falcon launches increased by 31 from 134 in 2024 to 165 in 2025, Space customer launches and average price

per launch remained relatively flat year over year.

110

Cost of Revenue

Cost of revenue for the year ended December 31, 2025decreased by $189 million,or 12.2%, compared to the prior

year ended December 31, 2024. This decrease was primarily due increased reusability of our Falcon launch vehicles

resulting in lower deprecation of $240 million, lowering the cost of each launch, and lower overhead costs of $11

million. The decrease is also due to the relative increase in Starlink satellite launches from 89 launches in 2024 to

122 launches in 2025, resulting in relatively more of our launch operations and overhead costs capitalized in our

Connectivity segment of $14 million.  This decrease was partially offset by an increase in inventory excess and

obsolescence reserves of $51 million mainly due to less demand on rocket vehicle and spacecraft parts as reusability

has increased. 

Research and Development

Research and development for the year ended December 31, 2025increased by $1,169 million,or 63.7%, compared

to the prior year ended December 31, 2024. This increase was primarily driven by higher production costs of $779

million, higher launch costs of $218 million, and higher engineering costs of $185 million, due to the accelerated

investment in development of the Starship vehicle and continued development of production and launch facilities to

support future Starship launches. 

Selling, General, and Administrative

Selling, general, and administrative for the year ended December 31, 2025decreased by $26 million,or 6.9%,

compared to the prior year ended December 31, 2024. This decrease was primarily due to lower allocated general

and administrative overhead of $52 million, partially offset by higher employee compensation expenses (including

salaries, benefits, and share-based compensation) of $16 million.

Impairment

Impairment for the year ended December 31, 2025increased by $14 million,or 61.5%, compared to the prior year

ended December 31, 2024. This increase was primarily due to a non-recurring impairment loss on a Falcon 9 booster

due to a post-landing anomaly during the year.

Income (Loss) from Operations

Space income from operations for the year ended December 31, 2025decreased by $678 million compared to the

prior year ended December 31, 2024 driven by the factors described above.

Connectivity

	Year Ended December 31,		2025 vs. 2024 Change
(in millions)	2025	2024	$ Change	% Change
Revenue...............................................................	$11,387	$7,599	$3,788	49.8%
Costs and expenses
Cost of revenue..............................................	5,921	4,768	1,153	24.2%
Research and development.............................	575	453	122	27.1%
Selling, general, and administrative...............	468	333	135	40.4%
Impairment.....................................................	—	39	(39)	NM
Total costs and expenses................................	$6,964	$5,593	$1,371	24.5%
Income from operations......................................	$4,423	$2,006	$2,417	120.4%

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

111

Revenue

Revenue for the year ended December 31, 2025increased by $3,788 million,or 49.8%, compared to the prior year

ended December 31, 2024. This increase was primarily driven by an increase of $2,377 million in revenue from our

consumer subscribers, composed of 99.9% growth in Starlink Subscribers, offset by an11.2%decline in Starlink

Subscriber ARPU, primarily due to international expansion and the addition of lower priced service plans. In

addition, Connectivity revenue had an increase of $1,411 million from our enterprise and government customers,

primarily driven by the growth in our enterprise connectivity business of $1,218 million inclusive of growth in our

mobile connectivity business of $632 million, and growth in our government connectivity business of $193 million.

Cost of Revenue

Cost of revenue for the year ended December 31, 2025increased by $1,153 million,or 24.2%, compared to the prior

year ended December 31, 2024. This increase was primarily due to higher depreciation of $827 million from

capitalized launch and satellite costs, higher operating expenses of $283 million mainly driven by ground operating

costs of $134 million, payment processor fees of $45 million, international expansion of $44 million, warranty costs

of $38 million, and employee compensation expenses (including salaries, benefits, and share-based compensation)

of $12 million, and higher freight costs of $72 million.

Research and Development

Research and development for the year ended December 31, 2025increased by $122 million,or 27.1%, compared to

the prior year ended December 31, 2024. This increase was primarily due to higher costs for the next-generation

production development of satellites of $84 million, Starlink Kits of $22 million, and ground equipment of $15

million.

Selling, General, and Administrative

Selling, general, and administrative for the year ended December 31, 2025increased by $135 million,or 40.4%,

compared to the prior year ended December 31, 2024. This increase was primarily driven by higher marketing costs

of $53 million, higher international expansion costs of $37 million, and higher allocated general and administrative

overhead of $67 million.

Impairment

Impairment for the year ended December 31, 2025decreased by $39 million compared to the prior year ended

December 31, 2024. The decrease was primarily related to the discontinuation of a Starlink Kit production line in

2024 with no impairment in 2025.

Income from Operations

Connectivity income from operations for the year ended December 31, 2025increased by $2,417 million,or

120.4%, compared to the prior year ended December 31, 2024 driven by the factors described above.

112

AI

	Year Ended December 31,		2025 vs. 2024 Change
(in millions)	2025	2024	$ Change	% Change
Revenue...............................................................	$3,201	$2,620	$581	22.2%
Costs and expenses
Cost of revenue..............................................	2,178	1,687	491	29.1%
Research and development.............................	5,064	1,176	3,888	330.8%
Selling, general, and administrative...............	1,827	1,105	722	65.4%
Restructuring charges.....................................	487	213	274	129.1%
Total costs and expenses...........................	$9,556	$4,181	$5,375	128.6%
Loss from operations...........................................	$(6,355)	$(1,561)	$(4,794)	307.1%

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the year ended December 31, 2025increased by $581 million,or 22.2%, compared to the prior year

ended December 31, 2024. This increase was primarily due to an increase in advertising revenue of $116 million as 

advertising spend increased from advertising partners on X and an increase in AI solutions and infrastructure

revenue of $465 million.  The increase in AI solutions and infrastructure revenue is mainly due to an increase in X

and Grok subscription revenue of $365 million and an increase in revenue from data licensing arrangements of $88

million.     

Cost of Revenue

Cost of revenue for the year ended December 31, 2025increased by $491 million,or 29.1%, compared to the prior

year ended December 31, 2024. This increase was primarily due to higher infrastructure and cloud computing costs

of $412 million attributable to increased subscriber revenue, higher employee compensation expenses (including

salaries, benefits, and share-based compensation) of $90 million, higher revenue share and content creator fees of

$45 million, and higher payment processor fees of $28 million, partially offset by a decrease in depreciation and

amortization expense of $97 million driven by a decrease in amortization expense for intangible assets that were

fully amortized during 2025.

Research and Development

Research and development for the year ended December 31, 2025increased by $3,888 million,or 330.8%,

compared to the prior year ended December 31, 2024. This increase was primarily due to higher GPU depreciation

expense of $1,673 million, higher infrastructure and cloud computing expenses of $1,440 million associated with the

build out of our compute infrastructure, and higher employee compensation expenses (including salaries, benefits,

and share-based compensation) and allocated overhead costs of $775 million.

Selling, General, and Administrative

Selling, general, and administrative for the year ended December 31, 2025increased by $722 million,or 65.4%,

compared to the prior year ended December 31, 2024. This increase was primarily due to higher employee

compensation expenses (including salaries, benefits, and share-based compensation) of $519 million as we continue

to expand our AI business, higher legal expenses of $189 million, and higher facilities and general and

administrative costs of $14 million.

Restructuring Charges

Restructuring charges for the year ended December 31, 2025increased by $274 million or 129.1%, compared to the

prior year ended December 31, 2024. This increase was primarily due to additional expense recorded to settle with

former Twitter employees as part of the workforce reduction program implemented in 2022.

113

Loss from Operations

AI loss from operations for the year ended December 31, 2025 increased by $4,794 million, or 307.1%, compared to

the prior year ended December 31, 2024 driven by the factors described above.

Comparison of the Years Ended December 31, 2024 and 2023

Consolidated Results of Operations

	Year Ended December 31,		2024 vs. 2023 Change
(in millions)	2024	2023	$ Change	% Change
Revenue...............................................................	$14,015	$10,387	$3,628	34.9%
Costs and expenses
Cost of revenue..............................................	7,996	6,110	1,886	30.9%
Research and development.............................	3,464	2,105	1,359	64.6%
Selling, general, and administrative...............	1,813	1,665	148	8.9%
Restructuring charges.....................................	213	237	(24)	(10.1)%
Impairment.....................................................	63	3,775	(3,712)	(98.3)%
Total costs and expenses...........................	13,549	13,892	(343)	(2.5)%
Income (loss) from operations............................	466	(3,505)	3,971	NM
Interest expense...................................................	(1,580)	(1,693)	113	(6.7)%
Interest income....................................................	371	249	122	49.0%
Other income, net................................................	985	(42)	1,027	NM
Income (loss) before income taxes......................	242	(4,991)	5,233	NM
Benefit from income taxes..................................	(549)	(363)	(186)	51.2%
Net income (loss)................................................	$791	$(4,628)	$5,419	NM

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the year ended December 31, 2024 increased by $3,628 million, or 34.9%, compared to the prior year

ended December 31, 2023. This increase was primarily due to an increase in revenue from our Connectivity segment

of $3,730 million as both our Starlink consumer subscriber base continued to grow as well as our Connectivity

enterprise and governmentsales, and an increase in revenue from our Space segment of $239 million due to the

increase in Falcon 9 launches partially offset by a decrease in Launch and Development revenue due to timing of

government contracts. This increase was partially offset by a decrease in revenue from our AI segment of $341

million driven by a decrease in advertising sales, partially offset by an increase in X subscriptions and data licensing

arrangements.

Cost of Revenue

Cost of revenue for the year ended December 31, 2024 increased by $1,886 million, or 30.9%, compared to the prior

year ended December 31, 2023. This increase was primarily due to a higher cost of revenue from the Connectivity

segment of $1,982 millionas a result of the higher volume spend on Starlink Kits as deliveries increased and higher

depreciation of launch costs driven by an increase in the number of satellites placed into orbit, partially offset by

cost efficiency from increased reusability of our Falcon launch vehicles in our Space segment of $128 million.

Research and Development

Research and development for the year ended December 31, 2024 increased by $1,359 million, or 64.6%, compared

to the prior year ended December 31, 2023. This increase was primarily due to higher cost in our AI segment of

114

$990 million related to advancing our AI technologies and higher costs of $297 million in our Space segment for

investment in Starship production, launch and engineering costs, and related facilities.

Selling, General, and Administrative

Selling, general, and administrative for the year ended December 31, 2024 increased by $148 million, or 8.9%,

compared to the prior year ended December 31, 2023.  This increase was primarily due to: (i) higher international

expansion costs of $18 million, higher employee compensation expenses (including salaries, benefits, and share-

based compensation) of $11 million, and higher allocated general and administrative overhead of $54 million in our

Connectivity segment, and  (ii) higher employee compensation expenses (including salaries, benefits, and share-

based compensation) and professional fees of $25 million in our Space segment.

Restructuring Charges

Restructuring charges for the year ended December 31, 2024 decreased by $24 million, or 10.1%, compared to the

prior year ended December 31, 2023.  This decrease was due to the impairment on office leases assumed as part of

the Twitter acquisition that occurred during the year ended December 31, 2023, partially offset by an increase in

workforce-related restructuring charges.

Impairment

Impairment for the year ended December 31, 2024 decreased by $3,712 million, or 98.3%, compared to the prior

year ended December 31, 2023. The impairment during the year ended December 31, 2023 was primarily related to

the impairment of the Twitter brand following its rebranding to X.

Income (Loss) from Operations

Income from operations for the year ended December 31, 2024 increased by $3,971 million compared to the prior

year ended December 31, 2023 driven by the factors described above.

Interest Expense

Interest expense for the year ended December 31, 2024 decreased by $113 million, or 6.7%, compared to the prior

year ended December 31, 2023. This decrease was primarily due to the debt issuance costs related to the X Bridge

Credit Facilities being amortized only through July 2024, the original maturity date, as compared to a full year of

amortization in 2023.

Interest Income

Interest income for the year ended December 31, 2024 increased by $122 million, or 49.0%, compared to the prior

year ended December 31, 2023. This increase was primarily due to an increase in dividend income earned from

marketable securities.

Other Income (Expense), net

Other income (expense), net for the year ended December 31, 2024 increased by $1,027 million compared to the

prior year ended December 31, 2023. This increase was primarily due to an unrealized gain on digital assets.

Benefit from Income Taxes

Benefit from income taxes for the year ended December 31, 2024 increased by $186 million,or 51.2%, compared to

the prior year ended December 31, 2023. This increase was primarily due to the change in the realizability of our net

deferred tax assets. As of December 31, 2024, we forecasted additional deferred tax assets related to U.S. R&D

credits would be utilized.

115

Net Income (Loss)

Net income for the year ended December 31, 2024 increased by $5,419 million compared to the prior year ended

December 31, 2023 driven by the factors described above.

Space

	Year Ended December 31,		2024 vs. 2023 Change
(in millions)	2024	2023	$ Change	% Change
Revenue...............................................................	$3,796	$3,557	$239	6.7%
Costs and expenses
Cost of revenue..............................................	1,541	1,669	(128)	(7.6)%
Research and development.............................	1,835	1,538	297	19.3%
Selling, general, and administrative...............	375	351	24	7.0%
Impairment.....................................................	24	—	24	NM
Total costs and expenses...........................	$3,775	$3,558	$217	6.1%
Income (loss) from operations............................	$21	$(1)	$22	NM

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the year ended December 31, 2024 increased by $239 million,or 6.7%, compared to the prior year

ended December 31, 2023. Launch Services revenue increased by $620 million as total Falcon launches increased by

38 from 96 in 2023 to 134 in 2024, with Launch Services missions increasing by 8.  This increase was partially

offset by a decrease of $381 million for Launch and Development revenue due to decreased activity in our

International Space Station contracts and lower revenue from a U.S. Department of War contract.

Cost of Revenue

Cost of revenue for the year ended December 31, 2024 decreased by $128 million,or 7.6%, compared to the prior

year ended December 31, 2023. This decrease was primarily due to increased reusability of our Falcon launch

vehicles resulting in lower depreciation of $80 million, lowering the cost of each launch. The decrease was also due

to the relative increase in Starlink satellite launches from 63 launches in 2023 to 89 launches in 2024, resulting in

relatively more of our launch operations and overhead costs capitalized in our Connectivity segment of $99 million. 

This decrease was offset by an increase in launch overhead costs of $77 million due to the increase in Falcon

launches. 

Research and Development

Research and development for the year ended December 31, 2024 increased by $297 million,or 19.3%, compared to

the prior year ended December 31, 2023. This increase was primarily due to higher production costs of $159 million,

higher launch costs of $67 million, and higher engineering costs of $56 million due to the increased investment in

the development of the Starship vehicle and related launch facilities.

Selling, General, and Administrative

Selling, general, and administrative for the year ended December 31, 2024 increased by $24 million,or 7.0%,

compared to the prior year ended December 31, 2023. This increase was primarily due to higher employee

compensation expenses (including salaries, benefits, and share-based compensation) and professional fees of $25

million.

116

Impairment

Impairment for the year ended December 31, 2024 increased by $24 million compared to the prior year ended

December 31, 2023. This increase was primarily due to non-recurring impairment losses resulting from one-time

launch anomalies experienced during the year.

Income (Loss) from Operations

Income (loss) from operations for the year ended December 31, 2024 increased by $22 million compared to the prior

year ended December 31, 2023 driven by the factors described above.

Connectivity

	Year Ended December 31,		2024 vs. 2023 Change
(in millions)	2024	2023	$ Change	% Change
Revenue...............................................................	$7,599	$3,869	$3,730	96.4%
Costs and expenses
Cost of revenue..............................................	4,768	2,786	1,982	71.1%
Research and development.............................	453	381	72	18.8%
Selling, general, and administrative...............	333	233	100	43.0%
Impairment.....................................................	39	—	39	NM
Total costs and expenses...........................	$5,593	$3,400	$2,193	64.5%
Income from operations......................................	$2,006	$469	$1,537	327.4%

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the year ended December 31, 2024 increased by $3,730 million,or 96.4%, compared to the prior year

ended December 31, 2023. This increase was primarily driven by an increase of$2,013 million in revenue from our

consumer subscribers, composed of96.5%growth in Starlink Subscribers offset by a 8.1% decline in Starlink

Subscriber ARPU primarily due to international expansion.  In addition, Connectivity revenue had an increase of

$1,717 million from our enterprise and government customers, primarily driven by the growth in our enterprise

connectivity business of $466 million and growth in our government connectivity business of $1,250 million.

Cost of Revenue

Cost of revenue for the year ended December 31, 2024 increased by $1,982 million,or 71.1%, compared to the prior

year ended December 31, 2023. This increase was primarily due to higher volume spend on Starlink Kits of $907

million driven by higher kit deliveries and higher depreciation of $555 million from capitalized launch and satellite

costs driven by an increase in the number of launches and satellites placed into orbit.

Research and Development

Research and development for the year ended December 31, 2024 increased by $72 million,or 18.8%, compared to

the prior year ended December 31, 2023. This increase was primarily due to higher costs for the next-generation

production development of satellites of $73 million, ground equipment of $4 million, offset by lower costs of $4

million for Starlink Kits.

Selling, General, and Administrative

Selling, general, and administrative for the year ended December 31, 2024 increased by $100 million,or 43.0%,

compared to the prior year ended December 31, 2023. This increase was primarily due to higher international

expansion costs of $18 million, higher employee compensation expenses (including salaries, benefits, and share-

based compensation) of $11 million, and higher allocated general and administrative overhead of $54 million.

117

Impairment

Impairment for the year ended December 31, 2024 increased by $39 million compared to the prior year ended

December 31, 2023. This increase was due to a discontinuation of a certain Starlink Kit production line.

Income from Operations

Income from operations for the year ended December 31, 2024 increased by $1,537 million, or 327.4%, compared to

the prior year ended December 31, 2023 driven by the factors described above.

AI

	Year Ended December 31,		2024 vs. 2023 Change
(in millions)	2024	2023	$ Change	% Change
Revenue...............................................................	$2,620	$2,961	$(341)	(11.5)%
Costs and expenses
Cost of revenue..............................................	1,687	1,655	32	1.9%
Research and development.............................	1,176	186	990	531.5%
Selling, general, and administrative...............	1,105	1,081	24	2.3%
Restructuring charges.....................................	213	237	(24)	(10.2)%
Impairment.....................................................	—	3,775	(3,775)	NM
Total costs and expenses...........................	$4,181	$6,934	$(2,753)	(39.7)%
Loss from operations...........................................	$(1,561)	$(3,973)	$2,412	(60.7)%

_________________

NM — Absolute percentage comparisons from positive to negative values or to zero values are considered not meaningful.

Revenue

Revenue for the year ended December 31, 2024 decreased by $341 million,or 11.5%, compared to the prior year

ended December 31, 2023. This decrease was due to a decrease in advertising revenue of $595 million, partially

offset by an increase in AI solutions and infrastructure revenue of $254 million. The decrease in advertising revenue

was due to the loss of advertising partners for X. The increase in AI solutions and infrastructure was due to an

increase in X subscription revenue of $157 million and an increase in data licensing arrangements of $90 million.  In

2023 and 2024, substantially all of our AI segment revenue consisted of advertising, subscriptions, and data

licensing revenue generated from X, formerly known as Twitter. 

Cost of Revenue

Cost of revenue for the year ended December 31, 2024 increased by $32 million,or 1.9%, compared to the prior

year ended December 31, 2023. This increase was primarily due to higher server depreciation of $97 million,

partially offset by lower infrastructure and revenue share expenses of $46 million, and lower employee and

facilities-related expenses of $18 million resulting from the Company’s restructuring and cost reduction efforts.

Research and Development

Research and development for the year ended December 31, 2024 increased by $990 million,or 531.5%, compared

to the prior year ended December 31, 2023. This increase was primarily due to increased investments made in

advancing our AI technologies, including employee compensation expenses (including salaries, benefits, and share-

based compensation) and infrastructure services of $703 million and higher depreciation of $321 million for our

equipment hardware.

Selling, General, and Administrative

Selling, general, and administrative for the year ended December 31, 2024 increased by $24 million,or 2.3%,

compared to the prior year ended December 31, 2023. This increase was primarily due to an increase in our

118

amortization expense of $107 million related to the Twitter brand becoming a finite-lived intangible asset and higher

legal costs of $65 million, partially offset by lower employee and facilities related costs of $125 million and lower

professional fees of $23 million resulting from the Company’s restructuring and cost reduction efforts.

Restructuring charges

Restructuring charges for the year ended December 31, 2024 decreased by $24 million,or 10.2%, compared to the

prior year ended December 31, 2023. This decrease was due to the impairment on the office leases assumed as part

of the Twitter acquisition that primarily occurred during the year ended December 31, 2023, partially offset by an

increase in workforce-related restructuring charges.

Impairment

Impairment for the year ended December 31, 2024 decreased by $3,775 million compared to the prior year ended

December 31, 2023. The impairment during the year ended December 31, 2023 was related to the impairment of the

Twitter brand intangible asset following its rebranding to X.

Loss from Operations

Loss from operations for the year ended December 31, 2024 decreased by $2,412 million, or 60.7%, compared to the

prior year ended December 31, 2023 driven by the factors described above.

Non-GAAP Financial Measures

Management believes that certain financial measures that are not presented in accordance with GAAP provide

management and investors with useful supplemental information that provides a meaningful view of our financial

condition and results of operations across periods by removing the impact of items that management believes do not

directly reflect our ongoing operating performance. Adjusted EBITDA and Segment Adjusted EBITDA are

supplemental measures that are not required by or presented in accordance with GAAP. In evaluating our

performance as measured by Adjusted EBITDA and Segment Adjusted EBITDA, management recognizes and

considers the limitations of these measures. Other companies in our industry may calculate Adjusted EBITDA and

Segment Adjusted EBITDA differently than we do or may not calculate them at all, limiting their usefulness as

comparative measures. Because of these limitations,  Adjusted EBITDA and Segment Adjusted EBITDA should not

be considered in isolation or as a substitute for net income (loss), income (loss) from operations, or any other

measure calculated in accordance with GAAP, and should be considered together with our GAAP financial

measures and the reconciliations to the corresponding most directly comparable GAAP financial measures set forth

in this prospectus.

Adjusted EBITDA is defined as net income (loss) excluding (i) depreciation and amortization, (ii) share-based

compensation, (iii) impairment, (iv) restructuring charges, (v) interest expense, (vi) interest income, (vii) other

income (expense), net and (viii) provision for income taxes. Segment Adjusted EBITDA is defined as segment

income (loss) from operations excluding (i) depreciation and amortization, (ii) share-based compensation, (iii)

restructuring charges, and (iv) impairment. Adjusted EBITDA and Segment Adjusted EBITDA are key performance

measures that our management uses to assess our financial performance as well as for internal planning and

forecasting purposes. We consider Adjusted EBITDA and Segment Adjusted EBITDA to be meaningful

performance measures for investors to evaluate our operating performance and to compare the financial results

between periods.

119

The following table sets forth a reconciliation of Net income (loss), the most directly comparable GAAP measure, to

Adjusted EBITDA:

	Three Months Ended March 31,		Year Ended December 31,
(in millions)	2026	2025	2025	2024	2023
Net income (loss)........................................	$(4,276)	$(528)	$(4,937)	$791	$(4,628)
Add (deduct):
Depreciation and amortization....................	2,442	1,443	6,701	3,824	2,635
Share-based compensation..........................	639	232	1,947	784	679
Restructuring charges..................................	(11)	4	487	213	237
Impairments................................................	—	24	38	63	3,775
Interest expense...........................................	664	447	1,945	1,580	1,693
Interest income............................................	(213)	(117)	(492)	(371)	(249)
Other (income) expense, net.......................	1,876	211	177	(985)	42
Provision for (benefit from) income taxes..	6	14	718	(549)	(363)
Adjusted EBITDA .....................................	$1,127	$1,730	$6,584	$5,350	$3,821

120

The following table sets forth a reconciliation of Income (loss) from operations for each segment, the most directly

comparable GAAP measure, to Segment Adjusted EBITDA:

	Three Months Ended March 31,
	2026
(in millions)	Space	Connectivity	AI	Total Reportable Segments
Income (loss) from operations............................	$(662)	$1,188	$(2,469)	$(1,943)
Add:
Depreciation and amortization............................	166	783	1,493	2,442
Share-based compensation..................................	145	116	378	639
Restructuring charges..........................................	—	—	(11)	(11)
Segment Adjusted EBITDA................................	$(351)	$2,087	$(609)	$1,127

	Three Months Ended March 31,
	2025
(in millions)	Space	Connectivity	AI	Total Reportable Segments
Income (loss) from operations............................	$(70)	$1,033	$(936)	$27
Add:
Depreciation and amortization............................	162	510	771	1,443
Share-based compensation..................................	108	75	49	232
Restructuring charges..........................................	—	—	4	4
Impairment..........................................................	24	—	—	24
Segment Adjusted EBITDA................................	$224	$1,618	$(112)	$1,730

	Year Ended December 31,
	2025
(in millions)	Space	Connectivity	AI	Total Reportable Segments
Income (loss) from operations............................	$(657)	$4,423	$(6,355)	$(2,589)
Add:
Depreciation and amortization............................	757	2,376	3,568	6,701
Share-based compensation..................................	515	369	1,063	1,947
Restructuring charges..........................................	—	—	487	487
Impairment..........................................................	38	—	—	38
Segment Adjusted EBITDA................................	$653	$7,168	$(1,237)	$6,584

121

	Year Ended December 31,
	2024
(in millions)	Space	Connectivity	AI	Total Reportable Segments
Income (loss) from operations............................	$21	$2,006	$(1,561)	$466
Add:
Depreciation and amortization............................	637	1,508	1,679	3,824
Share-based compensation..................................	472	296	16	784
Restructuring charges..........................................	—	—	213	213
Impairment..........................................................	24	39	—	63
Segment Adjusted EBITDA................................	$1,154	$3,849	$347	$5,350

	Year Ended December 31,
	2023
(in millions)	Space	Connectivity	AI	Total Reportable Segments
Income (loss) from operations............................	$(1)	$469	$(3,973)	$(3,505)
Add:
Depreciation and amortization............................	571	884	1,180	2,635
Share-based compensation..................................	427	249	3	679
Restructuring charges..........................................	—	—	237	237
Impairment..........................................................	—	—	3,775	3,775
Segment Adjusted EBITDA................................	$997	$1,602	$1,222	$3,821

Liquidity and Capital Resources

Our primary sources of liquidity are cash flows generated from operations, our total cash and cash equivalents of

$15,852 million as of March 31, 2026, short-term marketable securities of $7,823 millionas of March 31, 2026, and

borrowings under our credit facilities. As of March 31, 2026, we have $1,500 million available to borrow under the

SpaceX Credit Facility.  The cash we generate from our core operations also enables us to fund our research and

development projects including our Starship rocket and next-generation satellites, the construction of future data

centers, and the continued expansion of our AI-enabled products.

In addition, because we expect a significant portion of our future expenditures to fund growth initiatives, we retain

flexibility to adjust spending across segments. For example, if our near-term data center needs decrease in scale or

ramp more slowly than expected, including due to global economic, tax, trade or business conditions, we may

reduce future capital expenditures in this segment and reallocate those expenditures to other segments based on

business priorities and growth opportunities. In addition, we continually evaluate our cash needs and may decide it is

best to raise additional capital or seek alternative financing sources to fund the rapid growth of our business,

including through drawdowns on existing or new debt facilities. We may seek to refinance the SpaceX Bridge Loan,

including with the proceeds from notes offerings, bank borrowings, or other financial arrangements. We may also

from time to time determine that it is in our best interests to voluntarily repay certain indebtedness early.

Accordingly, we believe we have sufficient sources of funding to meet our business requirements for at least the

next twelve months from the issuance of the consolidated financial statements.

Debt Agreements

SpaceX Credit Facility

In February 2025, SpaceX entered into a five-year senior unsecured revolving credit agreement with a syndicate of

banks, under which the Company may borrow up to $1,500 million (“SpaceX Credit Facility”). The SpaceX Credit

Facility is subject to certain customary representations, warranties, covenants, and events of default, including a

122

maximum financial covenant requiring the Company to maintain a Consolidated Leverage Ratio (as defined in the

SpaceX Credit Facility) of no greater than 3.75 to 1.0 as of the end of each fiscal quarter (subject to temporary

increases to 4.25 to 1.0 following certain qualified acquisitions) and other customary reporting requirements. The

SpaceX Credit Facility also includes sublimits of up to $150 million for financial letters of credit and up to $1,000

million for performance letters of credit. The SpaceX Credit Facility terminates, and all outstanding loans become

due and payable, on February 7, 2030, unless the parties agree to an extension in accordance with the terms of the

SpaceX Credit Facility. As of March 31, 2026 and December 31, 2025, no amounts were outstanding under the

SpaceX Credit Facility.

Borrowings under the SpaceX Credit Facility bear interest, at the Company’s option, at a rate per annum equal to (i)

a forward-looking term rate based on SOFR (“Term SOFR”) plus an applicable margin ranging from 0.75% and

1.25% (depending on the Company’s debt rating), or (ii) a base rate equal to the highest of (a) Federal Funds Rate

plus 0.5%, (b) the Prime Rate, (c) Term SOFR plus 1.00%, and (d) 1.00% plus an applicable margin ranging from

0.0% and 0.25% (depending on the Company’s debt rating). The Company may also borrow in various alternative

currencies, with interest calculated at rates based on SONIA for Pound Sterling-denominated loans and EURIBOR

for Euro-denominated loans, plus an applicable margin. In addition, the Company pays a commitment fee on the

unused portion of the SpaceX Credit Facility, which ranges from 0.07% (amended to 0.06% under the Amended

SpaceX Credit Facility described below) to 0.11% per annum based on the Company’s debt rating. As of March 31,

2026, the Company was in compliance with all covenants under the SpaceX Credit Facility.

In March 2026, the Company entered into a First Amendment to Credit Agreement and Waiver (the “First

Amendment”) with its lenders, in connection with the Company’s entry into the SpaceX Bridge Loan (as defined

below). The First Amendment, among other things, (i) waived certain specified defaults and (ii) amended certain

definitions and covenants under the SpaceX Credit Facility to conform to the terms of the SpaceX Bridge Loan.

In May 2026, SpaceX amended the SpaceX Credit Facility to increase the borrowing capacity up to $5,000 million

(“Amended SpaceX Credit Facility”). As part of the Amended SpaceX Credit Facility, the sublimit for performance

letters of credit was increased to $2,000 million. The Amended SpaceX Credit Facility terminates, and all

outstanding loans become due and payable, on May 19, 2031, unless the parties agree to an extension in accordance

with the terms of the Amended SpaceX Credit Facility. All other terms were consistent with the terms of the SpaceX

Credit Facility.

SpaceX Bridge Loan

In March 2026, SpaceX entered into a new bridge loan credit agreement (the “SpaceX Bridge Loan”) with a

syndicate of lenders, providing for an unsecured bridge term loan facility in an aggregate principal amount of

$20,000 million. The SpaceX Bridge Loan matures on September 2, 2027, with two three-month extensions at the

Company’s option, subject to the absence of a continuing default and the payment of an extension fee of 0.25% of

the aggregate outstanding principal per extension, resulting in a final extended maturity date of March 2028.

The proceeds of the SpaceX Bridge Loan were used to repay the X B-1 Term Loan, the X B-3 Term Loan, the xAI

Fixed Rate Loan, the xAI Floating Rate Loan and the xAI 12.5% Senior Secured Notes (as defined and described in

Note 10, Debt, to the consolidated financial statements included elsewhere in this prospectus). The Company may

also use the remaining proceeds for general corporate purposes.

The SpaceX Bridge Loan bears interest, at the Company’s election, at a rate per annum equal to (i) Term SOFR plus

an applicable margin ranging from 0.75%-1.75% (depending on the Company’s debt rating), or (ii) a base rate equal

to the highest of (a) the Federal Funds Rate plus 0.5%, (b) the Prime Rate, (c) Term SOFR plus 1.0% and (d) 1.00%,

plus an applicable margin ranging from 0.00% to 0.75% (depending on the Company’s debt rating). In addition, the

Company is obligated to pay duration fees equal to 0.125% of outstanding principal on the first anniversary of

closing and 0.25% of outstanding principal on the fifteen-month anniversary of closing. As of March 31, 2026, the

Company was in compliance with all covenants under the SpaceX Bridge Loan.

The obligations of the Company under the SpaceX Bridge Loan are guaranteed on a joint and several basis by X

Corp., X.AI LLC, and CTC Property LLC (each a subsidiary of the Company). The SpaceX Bridge Loan may be

prepaid at any time, in whole or in part, without premium or penalty. The Company is required to use an amount

123

equal to the net cash proceeds of certain debt financings to repay amounts outstanding under the SpaceX Bridge

Loan and to apply an amount equal to the net proceeds of a qualified initial public offering, including this offering,

to repay such amounts within six months following receipt of such proceeds.

The SpaceX Bridge Loan contains customary events of default and affirmative and negative covenants, including

restrictions on liens, subsidiary indebtedness, fundamental changes (including a prohibition on the disposition of

Starlink assets and other material businesses outside the consolidated group), and changes in the nature of the

Company’s business. The sole financial maintenance covenant requires the Company to maintain a Consolidated

Leverage Ratio — defined as consolidated funded indebtedness (net of 85% of unrestricted cash) to Consolidated

EBITDA (as defined in the SpaceX Bridge Loan) — of no greater than 3.75 to 1.0 as of the end of each fiscal

quarter, with a temporary step-up to 4.25 to 1.0 for four fiscal quarters following a qualifying acquisition of at least

$1.0 billion.

Material Cash Commitments

From time to time in the ordinary course of business, we enter into agreements with suppliers for the purchase of

parts and raw materials to manufacture our products. However, due to contractual terms, variability in the precise

growth curves of our development and production ramps, and opportunities to renegotiate pricing, these contracts

generally do not have long-term binding and enforceable purchase orders, and the timing and magnitude of purchase

orders beyond the short term is difficult to accurately project. Because we do not have long-term purchase orders for

these parts and raw materials, future purchases may result in material cash commitments. For additional information

about this risk, please refer to “Risk Factors” in this prospectus.

On September 7, 2025, the Company entered into a License Purchase Agreement (the “Spectrum License Purchase

Agreement”) with Spectrum Business Trust 2025-1, a Nevada Business Trust (“Trust”) and EchoStar Corporation

(“EchoStar” and the transactions contemplated thereby, “Spectrum Transaction”). On November 5, 2025 the parties

amended and restated the Spectrum License Purchase Agreement to include EchoStar’s licenses for up to 15 MHz of

additional unpaired AWS-3 spectrum. The total consideration for the acquisition of EchoStar’s spectrum is

approximately $19.6 billion, consisting of (i) approximately $11.1 billion in equity, payable through the issuance of

approximately 261.8 million shares of the Company’s Class A common stock at a fixed value of $42.40 per share,

and (ii) up to $8.5 billion related to the payoff of designated EchoStar debt, with any shortfall below $8.5 billion to

be paid in cash. The allocation of cash and equity consideration is subject to certain adjustments based on the

amount of EchoStar debt satisfied at or prior to closing.The Spectrum Transaction was approved by the FCC on

May 12, 2026 and is expected to close on or about November 30, 2027 subject to other closing conditions. Upon

closing, the Company intends to either use cash and cash equivalents on hand or seek alternative financing sources

to fund the cash payment to EchoStar. 

As of March 31, 2026, we and our subsidiaries had outstanding$29,132 millionin aggregate principal amount of

indebtedness and no debt principal payments are due until August 28, 2027 if we choose not to extend.  As of March

31, 2026, our total minimum lease payments was $5,823 million, of which$1,026 million is due within this fiscal

year. For details regarding our indebtedness and lease obligations, refer to Note 10, Debt, and Note 11, Leases of our 

audited consolidated financial statements and Note 9, Debt of our unaudited consolidated financial statements

included elsewhere in this prospectus.

Summary of Cash flows

The following table summarizes our cash flows for the periods indicated:

	Three Months Ended March 31,		Year Ended December 31,
(in millions)	2026	2025	2025	2024	2023
Net cash provided by (used in)
Operating activities........................	$1,047	$727	$6,785	$5,776	$4,520
Investing activities.........................	$(16,724)	$(4,170)	$(19,575)	$(10,796)	$(4,867)
Financing activities........................	$7,125	$354	$26,350	$11,830	$422

124

Operating Activities

Net cash provided by operating activities increased by $320 million from $727 million during the three months

endedMarch 31, 2025 to $1,047 million during the three months endedMarch 31, 2026. This increase was primarily

driven by an increase in working capital for deferred revenue of $1,153 million from upfront payments from our

Space and Connectivity customers, partially offset by lower net income exclusive of non-cash items.

Net cash provided by operating activities increased by $1,009 million from $5,776 million during the year ended

December 31, 2024 to $6,785 million during the year ended December 31, 2025. This increase was primarily driven

by higher net income exclusive of non-cash items and an increase of $1,080 millionfor accounts payable and other

liabilities as we continue to expand our infrastructure and timing of payments, and higher deferred revenue from

cash received from upfront payments from our aviation customers. This increase was partially offset by an increase

of $449 million for accounts receivable, prepaid expenses, and inventory.

Net cash provided by operating activities increased by $1,256 million from $4,520 million during the year ended

December 31, 2023 to $5,776 million during the year ended December 31, 2024. This increase was primarily driven

by higher net income exclusive of non-cash items, partially offset by a decrease of $628 millionfor inventory,

accounts receivable, prepaid expenses and other assets due to increase in our revenue and production of Starlink

Kits.

Investing Activities

Net cash used in investing activities increased by $12,554 million from $4,170 million during the three months

endedMarch 31, 2025 to $16,724 million during the three months endedMarch 31, 2026.  This increase was

primarily driven by an increase in capital expenditures of $5,967 million related to the build out of data centers and

related infrastructure, and space launch facilities and related infrastructure, as well as an increase in purchases of

marketable securities of $7,489 million in the period. This increase was partially offset by an increase in cash

received from product rebates of $1,195 million.

Net cash used in investing activities increased by $8,779 million from $10,796 million during the year ended

December 31, 2024 to $19,575 million during the year ended December 31, 2025.  This increase was primarily

driven by an increase in capital expenditures of $9,574 million related to the build out of data centers and related

infrastructure, and space launch facilities and related infrastructure, partially offset by a net increase in cash received

from marketable securities of $1,264 million.

Net cash used in investing activities increased by $5,929 million from $4,867 million during the year ended

December 31, 2023 to $10,796 million during the year ended December 31, 2024.  This increase was primarily

driven by an increase in capital expenditures of $6,748 million related to the build out of data centers and related

infrastructure, and space launch facilities and related infrastructure, partially offset by an increase in cash received

for the maturities of marketable securities of $981 million.

Financing Activities

Net cash provided by financing activities increased by $6,771 million from $354 million during the three months

endedMarch 31, 2025 to $7,125 million during the three months endedMarch 31, 2026.  This increase was

primarily driven by an increase in proceeds from the SpaceX Bridge Loan and other financing arrangements of

$17,950 millionand proceeds from sale of our capital stock of $7,420 million, partially offset by an increase in

payment on existing debt obligations and debt extinguishment costs of $14,703 millionfrom the proceeds from the

SpaceX Bridge Loan as well as an increase in repurchases of our capital stock of $3,838 million following the xAI

Merger.

Net cash provided by financing activities increased by $14,520 million from $11,830 million during the year ended

December 31, 2024 to $26,350 million during the year ended December 31, 2025.  This increase was primarily

driven by an increase in proceeds from debt and other financing arrangements for our AI segment of $16,055 million

and proceeds from sale of our capital stock of $5,706 million, partially offset by an increase in repayments on debt

and other financing arrangements for our AI segment of $6,781 million.

125

Net cash provided by financing activities increased by $11,408 million from $422 million during the year ended

December 31, 2023 to $11,830 million during the year ended December 31, 2024.  This increase was primarily

driven by an increase in proceeds from the sale of our capital stock of $12,327 million, partially offset by an increase

in the buyback of common and preferred shares by the Company of $104 million.

Critical Accounting Estimates

The preparation of financial statements and related disclosures in conformity with GAAP and the Company’s

discussion and analysis of its financial condition and operating results require the Company’s management to make

judgments, assumptions and estimates that affect the amounts reported. Note 2, “Summary of Significant

Accounting Policies” of the Notes to audited consolidated financial statements included elsewhere in this prospectus

describes the significant accounting policies and methods used in the preparation of the Company’s consolidated

financial statements. Management bases its estimates on historical experience and on various other assumptions it

believes to be reasonable under the circumstances, the results of which form the basis for making judgments about

the carrying values of assets and liabilities.

Revenue Recognition

Space contract revenue is derived from fixed-price contracts related to the development and provision of launch

services for the deployment of spacecraft and other payloads to their intended orbit for both commercial customers

and governmental agency space programs. Connectivity contract revenue for Starshield customers is mostly derived

from fixed-price contracts related to the development of a secure satellite network designed specifically for

government and national security applications.

The Company recognizes revenue over time when the Company’s performance on the contract creates an asset with

no alternative use and when the Company has an enforceable right to payment for performance to date. The

Company measures progress on these contracts using the cost-to-cost input method, as the Company believes this

represents the most appropriate measure towards satisfaction of its performance obligation. Under the cost-to-cost

input method, the Company records revenue based upon costs (such as materials and labor hours) incurred to date

relative to the total estimated cost at completion.

The Company’s contracts recognized over time using the cost-to-cost input method are complex and require the

Company to estimate the total costs to perform over the term of the contracts, as well as the measurement of

progress towards completion for each performance obligation. For Space contracts, developing the estimated total

cost at completion for each performance obligation requires the use of significant management judgment, including

assumptions regarding launch timing, labor hours, allocation of shared costs for launch vehicles that have been

identified as reusable for multiple launches, as well as expected technological changes to launch vehicles and

spacecraft. For Connectivity contracts, developing the estimated total cost at completion for each performance

obligation requires the use of significant management judgment, including assumptions regarding labor hours,

allocation of shared costs used in the production of satellites, satellite material costs, as well as expected

technological changes to satellites. Material changes in estimated contract revenue or costs at completion and the

resulting changes in contract profit could have a material impact on the Company’s financial condition and operating

results.

The impact of net adjustments from contracts recognized over time using the cost-to-cost input method to our

revenue and operating income was not material for the years ended December 31, 2025, 2024, and 2023 and for the

three months ended March 31, 2026. If the combined gross margins for our contracts recognized over time using the

cost-to-cost input method had been estimated to be higher or lower by 1% during 2025, it would have increased or

decreased operating income for the year by approximately $110 million.

Property, Plant, and Equipment, Net

Property, plant, and equipment, net is stated at cost less accumulated depreciation. The Company depreciates these

assets primarily using the straight-line method over the estimated useful lives of the assets except flight vehicles and

spacecraft, which are depreciated over the expected number of average flights for each flight vehicle and spacecraft.

Leasehold improvements are depreciated over the shorter of their estimated useful lives or the related lease term.

126

Determining the useful lives and the number of average flights a flight vehicle and spacecraft can fly require the

Company to estimate the period over which we expect to recover the economic value of our property, plant, and

equipment. For each of our flight vehicle hardware and spacecraft, we consider recovery and refurbishment success

rates, refurbishment economics, customer acceptance limits that may prohibit the use of vehicles that have been

flown more than a certain number of launches, expected future launches included in the mission manifest, as well as

any anticipated retirement timing of certain flight vehicle and spacecraft models such as Falcon as a result of

anticipated transition to Starship to determine the expected number of average flights for each vehicle.

For our satellites assets, we consider factors such as on-orbit performance, orbit-raise timing, expected service

capability, and the evolution of constellation density and technology.

When we determine that the useful lives or expected remaining flights of assets are shorter or longer than we had

originally estimated, we adjust the rate of depreciation to reflect the assets' revised useful lives or number of

remaining flights.

The Company periodically evaluates impairment of its property, plant, and equipment assets whenever events or

circumstances indicate that the carrying value of an asset or asset group may not be recoverable. Factors we consider

to identify indicators of potential impairment include significant changes or planned changes in our use of certain

property, plant and equipment, technological developments that reduce the utility of the existing assets, declines in

forecasted cash flows, and significant negative industry or economic trends.

Impairment is assessed at the lowest level for which identifiable cash flows are largely independent of the cash flows

of other assets and liabilities. If estimated future cash flows are less than the carrying value of the asset or asset

group, an impairment charge is recognized to the extent its carrying value exceeds its estimated fair value to cost of

revenue or selling, general, and administrative expenses depending on the nature of the assets, or to impairment

charges if the impairment is considered to be outside the normal course of business. For the years ended December

31, 2025, 2024, and 2023, and for the three months ended March 31, 2026, impairments on fixed assets were not

material.

If the average remaining flights for our flight vehicle and spacecraft had been estimated to be five more or fewer

flights, the impact to our operating income for the year ended December 31, 2025 and three months ended March 31,

2026 would not be material. If the average useful life of our satellite assets had been changed by one year, it would

have an approximately $480 million and $170 million impact on our operating income for the year ended December

31, 2025 and three months ended March 31, 2026, respectively.

Legal and Other Contingencies

The Company is subject to various legal proceedings and claims that arise in the ordinary course of business, the

outcomes of which are inherently uncertain. The Company records a liability when it is probable a loss has been

incurred and the amount is reasonably estimable, the determination of which requires significant judgment.

Resolution of legal matters in a manner inconsistent with management’s expectations could have a material impact

on the Company’s financial condition and operating results.

Recent Accounting Pronouncements

Refer toNote 2, Summary of Significant Accounting Policies, to the audited consolidated financial statements

included elsewhere in this prospectus.

Quantitative and Qualitative Disclosures About Market Risk

Foreign Currency Risk

Our Connectivity and AI businesses operate in many countries and transact in multiple currencies. In general, we are

a net receiver of currencies other than the U.S. dollar for our foreign subsidiaries. Accordingly, we are exposed to

foreign currency risk both from fluctuations in exchange rates affecting foreign-currency denominated transactions

and from the impact of translating the assets, liabilities, revenues, costs of revenue, and other operating expenses of

our foreign subsidiaries into U.S. dollars. We have experienced, and will continue to experience, fluctuations in our

127

net income as a result of gains (losses) on the settlement and the re-measurement of monetary assets and liabilities

not denominated in our functional currencies. We do not hedge foreign currency risk and changes in exchange rates

could have an adverse impact on our operating results and cash flows.

We considered the historical trends in foreign currency exchange rates and determined that it is reasonably possible

that adverse changes in foreign currency exchange rates of 10% for all currencies could be experienced in the near-

term. These changes were applied to our total monetary assets and liabilities denominated in our non-functional

currencies at the balance sheet date to compute the impact these changes would have had on our income (loss)

before income taxes. These changes would have resulted in an immaterial gain or lossas of March 31, 2026 and

December 31, 2025, respectively.

Interest Rate Risk

Our exposure to changes in interest rates relates primarily to our investment portfolio, interest income on cash and

cash equivalents and our credit facilities. 

Our cash and cash equivalents consist of cash, time deposits, money market funds, U.S. government and agency

securities. Our investment policy and strategy are focused on preservation of capital and supporting our liquidity

requirements. Changes in U.S. interest rates affect the interest earned on our cash and cash equivalents.  A

hypothetical 100 basis point increase or decrease in market interest rates would have resulted in an immaterial

increase or decrease in interest income for the year ended December 31, 2025 and three months ended March 31,

2026. 

The effective interest rate on outstanding borrowings under the SpaceX Bridge Loan was 4.58% as of March 31,

2026. A hypothetical 100 basis point increase in U.S. interest rates would increase annual interest expense by

approximately $200 million.

130

BUSINESS

“You want to wake up in the morning and think the future is going to be great—and that’s what being a space-faring

civilization is all about. It’s about believing in the future and thinking that the future will be better than the past. And

I can’t think of anything more exciting than going out there and being among the stars.”

—Elon Musk

Our Mission

Our mission is to build the systems and technologies necessary to make life multiplanetary, to understand the true

nature of the universe, and to extend the light of consciousness to the stars. To do this, we have formed the most

ambitious, vertically integrated innovation engine on (and off) Earth with unmatched capabilities to rapidly

manufacture and launch space-based communications that connect the world, to harness the Sun to power a truth-

seeking artificial intelligence that advances scientific discovery, and ultimately to build a base on the Moon and

cities on other planets.

Overview

Founded in 2002, SpaceX is the only company building the integrated hardware and software infrastructure of the

future across space, connectivity, and AI. At our core, we are builders. We design, manufacture, launch, and operate

products and services built on cutting-edge technologies, including the world’s most advanced rockets and

spacecraft. We safely and reliably transport astronauts, satellites, and other payloads on missions that benefit life on

Earth. Since 2023, we have launched more than 80% of mass to orbit for the world each year with an over 99%

mission success rate with Falcon rockets. We also operate a high-speed, low-latency global broadband data and

communications network powered by approximately 9,600 Starlink broadband and mobile satellites in Low-Earth

Orbit, delivering connectivity to millions of consumer, enterprise, and government customers across 164 countries,

territories, and other markets, as of March 31, 2026. Using our dedicated satellite-to-mobile constellation, we offer

connectivity services, supplementing terrestrial networks and substantially reducing mobile “dead zones” across

approximately 30 countries.

With the potential to improve both space exploration and life on Earth, AI accelerates SpaceX’s mission to make life

multiplanetary, to understand the true nature of the universe, and to extend the light of consciousness to the stars.

xAI, which was founded in 2023 and acquired by SpaceX in early 2026, is now an integral pillar of our vertically

integrated company. We are rapidly constructing AI compute infrastructure—starting on Earth with the goal of

extending to space—at industry-leading pace and cost efficiency. Our infrastructure supports training and inference

for Grok, which has emerged as one of the world’s most advanced frontier models. Grok is designed as a truth-

seeking AI model, built on our founder Elon Musk’s mission to enable humanity to understand the universe. We

believe that accomplishing this mission requires a truth-seeking approach to AI. We define truth seeking as the

active, relentless pursuit of what is objectively true about reality, and grounded in evidence, logic, empirical data,

and first principles thinking. Our goal is to understand and explain what the universe appears to be doing, as

accurately as current knowledge allows. Within two years of its initial model release, Grok achieved frontier-level

performance in scientific reasoning, as measured by its GPQA Diamond score, an industry benchmark that evaluates

AI models on a standardized set of questions written and validated by experts, on a faster timeline than reported by

other leading model providers. Grok also benefits from integration with X, our real-time information, entertainment,

and free speech platform, which serves as a foundational distribution and data engine for our AI ecosystem and

further enhances Grok’s truth-seeking objective.

We believe that space represents the largest economic frontier in human history, unlocking unprecedented

opportunities in orbit and on Earth. Earth has limits, so we must build infrastructure and industries in space,

expanding human capabilities to improve life on Earth and to establish life beyond. Connectivity infrastructure in

space is designed to help everyone on Earth have access to education, healthcare, entertainment, and

communications, and to enable people to overcome many traditional limits, such as physical and political borders.

We believe AI infrastructure in space can utilize the virtually limitless power of the Sun and thereby enable the use

of AI as a transformative force for understanding the universe and improving the daily lives of all humans. We

believe the convergence of these areas will enable an unprecedented expansion in the global economy, leading to an

131

age of abundance. Our innovations and technological advancements are redefining industries on Earth, while we aim

to create new ones on the Moon, Mars, and beyond. We are truly building the infrastructure of the future.

SpaceX is the only company that has cracked the code on accessing space at scale, revolutionizing an industry

characterized by decades of stagnation, risk aversion, and economically perverse cost structures. SpaceX upended

this paradigm through the application of first-principles thinking, which rejects industry assumptions and builds

solutions based on the fundamental laws of physics. Our intense, mission-driven, engineering-first culture and focus

on extreme vertical integration have propelled us to achieve what many deemed impossible. We have demonstrated

the ability to achieve groundbreaking technological innovations with speed, quality control, and precision. We

pioneered high-cadence, reliable, and affordable access to space with our Falcon family of rockets, with a goal to

transform the rocket launch industry into airline-like operations. In 2015, we established at least a 10-year lead over

the industry by successfully landing our first Falcon 9 booster back from space before anyone else. We have

continued to invest significantly in further increasing our lead by pursuing full and rapid reusability at scale,

including investing over $15 billion in our next-generation rocket, Starship.

We believe rocket launches and landings should be as routine and commonplace as airplanes taking off and landing.

To achieve this sort of cadence, our iterative approach emphasizes rapid designing, testing, and process

optimization, putting flight hardware in the flight environment as often as possible. This allows us to accelerate our

learning by repeatedly using and improving our systems. This has resulted in a significantly higher flight rate at

costs that are much lower than launch programs that existed before SpaceX. For example, according to NASA, the

first version of Falcon 9 in 2010 had a launch cost of approximately $2,700 per kilogram, which represented a

reduction of approximately 85% compared to the historical average launch cost per kilogram of $18,500. The first

version of Falcon Heavy in 2018 further reduced this cost to approximately $1,400 per kilogram, a reduction of

approximately 92% compared to the historical average cost. With the future deployment of Starship, which is

designed to be the world’s first fully and rapidly reusable spacecraft, we aim to further reduce the cost to reach orbit

by 99% or more relative to the historical average launch cost. Central to our cost advantage is the reusability of key

hardware—most notably boosters—which we recover, refurbish, and refly many times instead of discarding after

single use. This dramatically lowers per-launch costs by minimizing hardware replacement expenses and spreading

fixed production costs across repeated uses. Space flight that historically cost billions per launch now costs in the

tens of millions, fundamentally reducing the cost of space access, providing the opportunity to build new enterprises

in space.

Similarly, xAI has cracked the code in the complexities of building and scaling AI compute infrastructure, becoming

the first company to deploy a coherent gigawatt-scale AI training cluster. We believe the combination of our

proprietary AI infrastructure capability, our truth-seeking frontier model, Grok, and our access to real-time data on

X creates a formidable competitive advantage, allowing us to maintain a leading position in the development of

advanced artificial intelligence. This advantage stems from our complete vertical integration and the common vision

infused by our founder, Elon Musk. In just a few years, we have demonstrated an ability to build coherent compute

at scale and rapid speed with lower cost. COLOSSUS and COLOSSUS II collectively provide approximately 1.0

gigawatt of compute power, with additional power capacity available for data center operations. We believe speed is

a competitive advantage. In order to bring compute clusters online as fast as possible, we employ a vertically

integrated, nimble approach to construction. At COLOSSUS, we brought online the first cluster of approximately

100,000 H100 processors, approximately 130 megawatts of compute power, in just 122 days, repurposing the shell

of an existing factory. At COLOSSUS II, we brought online the first cluster of approximately 110,000 GB200

processors, approximately 210 megawatts of compute power, even faster in 91 days. As an illustrative comparison,

an industry benchmark to bring online a 100 megawatt greenfield data center is approximately two years.

Furthermore, in the case of COLOSSUS II, following the initial cluster, we brought online the second cluster of

110,000 GB300 processors and 220 megawatts of compute power in 64 days, demonstrating our ability to rapidly

scale our facilities once built. We expect that once fully operational, the next phase of expansion at COLOSSUS II

will bring online at least 220,000 additional GB300 processors and over 400 additional megawatts of compute

power. We also demonstrated asignificant improvement in cost efficiency, achieving data center construction costs

for COLOSSUS II that are considerably lower than industry benchmarks on a per megawatt basis.

We are able to deploy power and compute significantly faster than other AI companies through first-principles

thinking, behind-the-meter power generation, coupled with what we believe is the world’s largest network of

132

sustainable battery storage systems, and innovations in advanced liquid cooling, high-density rack layouts, and

efficient networking. Our facilities also incorporate innovative design features that limit the effects on regional

electricity pricing for neighbors and include advanced water cleaning, reclamation, and recycling processes to

support sustainable operations. We partner with utilities and communities to connect to and enhance the grid over

time, and do so while pledging to cover costs of all new power delivery infrastructure upgrades to service our data

centers, including adequate network upgrade costs, to ensure that these expenses are not passed on to the ordinary

household. Our ability to rapidly and cost-effectively scale with the latest processors keeps us ahead of competitors

who deploy traditional and more expensive methods. As a result, we believe COLOSSUS II became one of the

world’s first data centers to deploy GB200s and GB300s, the most advanced AI processors available at the time, at

significant scale, and is currently powering training for our next frontier models, including Grok-5. Furthermore,

through our Terafab initiative together with Tesla to build a manufacturing facility capable of producing 1 terawatt

per year of compute hardware, we intend to further extend our vertical integration to chip design and manufacturing

to alleviate potential future chip shortages at SpaceX, optimize compute performance, and potentially reduce overall

compute costs. Intel, which has the ability to design, fabricate, and package ultra-high-performance chips at scale,

also joined the Terafab project in early April 2026. Our shovels-to-tokens approach allows us to train and iterate our

frontier models at high velocity, accelerating development cycles, eliminating external bottlenecks, and driving

rapid, continuous improvements in model performance.

In pursuing our mission, SpaceX has created new opportunities across our three foundational competitive

advantages, Space, Connectivity, and AI:

•Space. Launch is one of our foundational competitive advantages. We were the first private company to

develop and launch a liquid-fuel rocket to reach orbit (2008), the first private company to successfully dock a

private spacecraft with the International Space Station (2012), the first company to propulsively land (2015) and

refly an orbital-class rocket booster (2017), the first to begin deploying a large-scale LEO broadband satellite

constellation (2019), and the first private company to launch astronauts to orbit, allowing American astronauts

to again fly to and from the International Space Station on an American launch vehicle(2020). As of March 31,

2026, SpaceX had completed approximately 650 orbital space launches, and over 540 of those launches were

completed by a flight-proven Falcon rocket, drastically reducing the cost of access to space. We are the only

private company that is certified by NASA to send human missions to orbit. We are currently developing

Starship, designed to be the world’s most powerful launch vehicle. Starship is designed to be a fully and rapidly

reusable transportation system capable of carrying larger payloads farther and at lower marginal cost per launch

than our current Falcon rockets. Our unparalleled launch capabilities power every aspect of our business.

•Connectivity. Since activating service for customers in 2020, Starlink has rapidly expanded global access to

high-speed internet, prioritizing underserved rural and remote communities worldwide. While building

terrestrial networks in such communities can be prohibitively expensive, Starlink is capable of delivering

broadband connectivity anywhere on Earth with just a Starlink Kit. As of March 31, 2026, we had

approximately 9,600 Starlink broadband and mobile satellites in Low-Earth Orbit, operating the world’s most

advanced broadband constellation providing internet connectivity to approximately 10.3 million Starlink

Subscribers across 164 countries, territories, and other markets. In January 2024, we also began deploying our

Starlink Mobile constellation that utilizes separate Starlink satellites with satellite-to-mobile capabilities,

substantially reducing mobile “dead zones” around the world. As of March 31, 2026, our dedicated satellite-to-

mobile constellation of approximately 650 V1 Mobile satellites provides satellite-to-mobile data, over-the-top

voice, and messaging services to approximately 7.4 million monthly unique devices across approximately 30

countries.

•AI.We were the first company to deploy a coherent gigawatt-scale AI training cluster. We own and operate

what we believe to be the largest AI training data center clusters on Earth, consisting of hundreds of thousands

GPUs—all in the same spirit that enabled us to launch Grok faster than any other leading foundational AI model

—while maintaining full vertical integration from on-site power generation and water reclamation to GPU

deployment. In under two years, we have established a dual advantage in both cost efficiency and deployment

speed at scale. By owning the compute infrastructure and vertically integrating across the full AI stack, we can

train and iterate our frontier models at lower cost and higher velocity and accelerate development cycles. This

eliminates external bottlenecks and drives rapid, continuous improvements in model performance. The addition

133

of the Terafab initiative aims to further extend our control to the foundational processor layer. We believe that

the key constraints in the continued growth of AI are physical—chip manufacturing, data center infrastructure,

and power generation; the future of AI will be determined by the control of the physical stack. We believe no

other AI company has better control over the full physical stack than SpaceX. We believe this combination of

our state-of-the-art AI compute infrastructure, our truth-seeking frontier model, and our access to real-time data

on X creates a significant strategic advantage. Our integrated AI platforms across Grok and X have over 1.3

billion supported accounts active in the last twelve months ended March 31, 2026, including approximately 550

million MAUs, up from over 1.1 billion supported accounts and approximately 520 million MAUs as of

December 31, 2025, and generating approximately 350 million daily posts. Of our MAUs, we had

approximately 117 million MAUs that used Grok’s AI features as of March 31, 2026. Grok’s deep integration

with X enables freshness, relevance, and contextual awareness that we believe is a competitive differentiator.

This direct, real-time access to the information and human discourse on X enhances Grok’s truth-seeking

capabilities by grounding outputs in up-to-date knowledge and diverse viewpoints. As a result, we believe Grok

can deliver the most objective and relevant insights and best serve high-frequency, high-value use cases across

consumer and enterprise AI applications.

For complex reasoning and agentic workloads, compute is directly correlated with the quality of intelligence

and task completion speed. Over the long-term, however, we expect Earth’s finite resources will not be able to

sustain the immense computational demands of advanced AI models. Sustainably satisfying this compute

demand will require space-based infrastructure that utilizes the ultimate fusion energy source: the Sun. We

believe we are the only company with a commercially viable path to building orbital AI compute at scale, due to

our unique ability to launch substantial mass into orbit through reusable, cost-efficient rockets, to manufacture

secure, reliable, and high-performance satellites at low cost and high volume, and to manage large-scale

constellations. We expect that owning scalable, power-efficient infrastructure to train and operate frontier

models will be the most important driver for AI differentiation as AI systems converge toward artificial general

intelligence (“AGI”)—which has the potential to unlock large-scale productivity gains, scientific discovery, and

societal abundance.

We have created distinct new markets across the space, connectivity, and AI industries by building the integrated

hardware and software infrastructure of the future and by combining our broad range of capabilities. For example,

SpaceX’s recent acquisition of xAI unites SpaceX’s launch capabilities and global connectivity network with xAI’s

AI development capabilities. Specifically, we believe SpaceX’s reusable rockets, scaled satellite manufacturing, and

operational expertise can enable the cost-effective and rapid deployment of massive AI compute satellite

constellations—with potentially millions of satellites—for orbital data centers. We believe these AI compute

satellites in Sun-synchronous orbit will be able to handle energy-intensive AI workloads, such as inference demand,

at far greater scale and efficiency than terrestrial alternatives, with Starlink providing low-latency, global

connectivity linking these orbital AI systems to people around the world and delivering real-time intelligence. Our

goal is to leverage our launch leadership, global connectivity network, and AI expertise to allow us to continue

building the integrated infrastructure of the future on Earth, the Moon, Mars, and beyond to benefit humanity.

135

We have an intense, mission-driven, engineering-first culture that seeks to achieve what many have deemed

impossible. “The Algorithm,” as it is known internally, is a five-step iterative process that emphasizes making the

requirements less dumb, deleting unnecessary processes or parts (embracing the principle that the best part is no

part), only then optimizing the necessary processes or parts, accelerating cycle time, and automating only proven

processes. We strive to make the incredible and extraordinary accessible and repeatable, and we have grown rapidly

by continuously leveraging our core strengths, including:

•Global leadership in orbital launch services;

•Unrivaled satellite and connectivity platform across design, manufacturing, deployment, and operations;

•Truth-seeking AI model enhanced by real-time data;

•Extreme vertical integration enabling high velocity and superior cost efficiency at scale;

•Unique ability to scale new trillion-dollar markets across Space, Connectivity, and AI;

•Business models that are incredibly difficult to replicate; and

•Our mission-driven culture and world-class talent.

We have a stellar track record of capital allocation and value creation in Space and Connectivity. Since SpaceX’s

founding in 2002, we have raised over $9 billion of equity capital to fund the development and growth of these two

business segments. The Space segment became Segment Adjusted EBITDA positive on a sustained basis beginning

in 2018 and the Connectivity segment became in aggregate Segment Adjusted EBITDA positive on a sustained basis

beginning in 2023. In 2025, our Space segment generated a loss from operations of $(657) million and Segment

Adjusted EBITDA of $653 million, including the impact of funding $3,004 million in research and development

expense for our next-generation Starship launch vehicle program. In 2025, our Connectivity segment generated

income from operations of $4,423 million and Segment Adjusted EBITDA of $7,168 million.

Our financial results reflect the strength of our operating model and our ability to create and scale multiple new

businesses:

•For the three months ended March 31, 2026, we generated revenue on a consolidated basis of $4,694 million,

loss from operations of $(1,943) million and Adjusted EBITDA of $1,127 million. In 2025, we generated

revenue on a consolidated basis of $18,674 million, loss from operations of $(2,589) million and Adjusted

EBITDA of $6,584 million. Our Space and Connectivity segments contributed the substantial majority of our

consolidated revenue in the three months ended March 31, 2026 and the year ended December 31, 2025,

demonstrating the benefits of their scale and operating leverage in our vertically integrated business model;

•For the three months ended March 31, 2026, our Space segment generated revenue of $619 million, loss from

operations of $(662) million, and Segment Adjusted EBITDA of $(351) million. In 2025, our Space segment

generated revenue of $4,086 million, loss from operations of $(657) million, and Segment Adjusted EBITDA of

$653 million Additionally, our Space segment funded $930 million and $3,004 million in research and

development expense during the three months ended March 31, 2026 and the year ended December 31, 2025,

respectively, for our next-generation Starship launch vehicle program. Starship is designed to enable a step-

function change in our launch capability across reusability, payload capacity, and launch cadence, and is the key

enabler of our long-term growth strategy by unlocking entirely new categories of missions;

•For the three months ended March 31, 2026, our Connectivity segment generated revenue of $3,257 million,

income from operations of $1,188 million, and Segment Adjusted EBITDA of $2,087 million. Our Connectivity

segment, primarily driven by Starlink, generated revenue of $11,387 million, income from operations of $4,423

million, and Segment Adjusted EBITDA of $7,168 million in 2025, representing year-over-year growth of

49.8%, 120.4%, and 86.2%, respectively, benefiting from subscriber growth, increasing enterprise adoption, and

continued improvement in network efficiency;

136

•In our newly acquired AI segment, we plan to prioritize growth and investment to capture significant

opportunities in AI applications and compute infrastructure. For the three months ended March 31, 2026, our AI

segment generated revenue of $818 million, loss from operations of $(2,469) million, and Segment Adjusted

EBITDA of $(609) million. In 2025, our AI segment generated revenue of $3,201 million, loss from operations

of $(6,355) million, and Segment Adjusted EBITDA of $(1,237) million, reflecting its earlier stage of

development and continued investments to support long-term growth opportunities in AI; and

•For the three months ended March 31, 2026, capital expenditures for our Space segment was $1,052 million, for

our Connectivity segment was $1,332 million and for our AI segment was $7,723 million. In 2025, capital

expenditures for our Space segment was $3,832 million, for our Connectivity segment was $4,178 million and

for our AI segment was $12,727 million. 

Segment Adjusted EBITDA is a non-GAAP measure. Please refer to the section titled “Management’s Discussion

and Analysis of Financial Condition and Results of Operations—Non-GAAP Financial Measures” for additional

information on our non-GAAP financial measures, including reconciliations of Segment Adjusted EBITDA to

segment income (loss) from operations, the most directly comparable GAAP measure.

Why This Matters Now

For the entirety of its existence, human civilization has lived on a single celestial body: Earth. The current paradigm,

in which human civilization is confined to one planet, exposes humanity to existential threats that are unpredictable

and uncontrollable on a planetary scale. These threats include naturally occurring catastrophic events—such as

asteroid impacts, volcanic activity, or solar fluctuations—as well as man-made global conflicts. Geological and

astronomical records indicate a non-zero probability of extinction-level events occurring over periods measurable in

millions of years. Reliance on a single planetary home constitutes a single point of failure and carries existential risk

with a probability of one that must be solved. By moving beyond the only home we have ever known, we ensure

species-level redundancy and that the light of consciousness will not be tied to a single planet subject to the

inevitable hazards of a harsh and vast universe. We do not want humans to have the same fate as dinosaurs. We want

to give them a reason to look ahead with excitement, with the prospect that we are entering an age of abundance

with an endlessly prosperous and exciting future.

Artist Visualization of Life on Mars

137

For decades, a reality where humanity travels between the planets and the stars has felt tantalizingly close but still

locked in the pages and screens of science fiction. We are capable of better understanding the universe, exploring the

universe, and ultimately making life multiplanetary across the universe. We are becoming a civilization with the

ability to reach beyond Earth’s cradle and begin to inhabit other worlds. While we remain dedicated to this

fundamental mission, our progress in accessing space continues to yield opportunities that enrich life on Earth.

We believe our steps into the expanse will be accelerated by the rapid emergence of AI. As humanity moves into the

unknown, we believe AI will be our greatest tool for innovation and navigation, helping us better understand day-to-

day life and the universe, and master the complexity of establishing new civilizations in the far-flung reaches of

space. For AI to help us understand the universe, we believe it must be able to discard the often popular, but wrong,

in favor of the unpopular, but true. By combining the innate human desire to seek truth and explore with our

breakthrough technologies, we believe humanity will eventually reach new frontiers across the universe, while

enhancing the quality and resilience of life on Earth.

The rapid emergence of the AI era intensifies the urgency of our mission, as AI has the potential to accelerate not

only space exploration, but also transformative societal advancements on Earth. However, AI’s ability to

revolutionize human potential is directly dependent on meeting exponentially increasing resource demands. On

Earth, the massive expansion of data center capacity to support growing compute demand is significantly outpacing

electricity generation, which was effectively flat in the United States for approximately 15 years, growing at a

compound annual growth rate of 0.1% from 2008 to 2023. Despite the recent increase in electricity demand from AI

data centers, electricity generation in the United States has grown at an annual rate of less than 3% between 2023

and 2025, while electricity generation in China has grown at approximately twice that rate in the same time period.

U.S. compute demand has already outpaced available power supply with estimated demand of 62 gigawatts in 2025

exceeding the power generation of 49 gigawatts, according to industry sources. We expect the gap between demand

for compute and power supply to continue to widen meaningfully as AI compute needs proliferate. Such structural

power shortages are expected to intensify over the coming years. This supply and demand imbalance is already

imposing unsustainable strains on terrestrial power grids, supply chains, and the environment. The Sun contains

approximately 99.8% of the solar system’s energy and, as a result, we believe it is the only truly scalable solution to

terrestrial energy constraints in the age of AI. Harnessing this energy in space is considerably more efficient than on

land. Space-based solar arrays can generate more than five times the energy per unit area of terrestrial solar due to

continuous illumination, lack of atmospheric interference, and optimal orientation. SpaceX is well-positioned to

capture this space-based solar energy through our ability to rapidly access Sun-synchronous orbit through our

satellite manufacturing scale and launch capability. As a result, we are expanding our footprint and harnessing the

vast resources of space that are essential to sustaining technological development. Our goal is to ensure that AI

becomes a force for human flourishing and a benefit to civilization, rather than a catalyst for terrestrial resource

depletion and instability. We believe owning scalable, power-efficient infrastructure to train and operate frontier

models will be the most important competitive differentiator as AI systems converge toward AGI—which has the

potential to unlock large-scale productivity gains, scientific discovery, and societal abundance.

We believe space represents the largest economic frontier in human history. Our unmatched launch cadence has

massively increased access to space, enabling rapid and reliable missions for humans, cargo, and satellites—creating

unprecedented opportunities for innovation, scientific discovery, and global connectivity. SpaceX has always been a

mission-driven company, founded with the goal of making humanity multiplanetary. By dramatically reducing the

cost of access to space, we have been able to expand our mission to address some of the Earth’s most pressing

challenges, including bridging the digital divide by aiming to connect over three billion unconnected people to the

internet and humanity’s collective knowledge. Starlink is our groundbreaking solution for global internet

connectivity, delivering high-speed, low-latency access to the most remote and underserved corners of the world—

from Antarctica’s frozen wilderness to vast oceans and towering mountaintops—overcoming barriers posed by

traditional terrestrial infrastructure. Starlink’s unparalleled global reach has the potential to enable society to educate

billions of people, to help lift entire communities out of poverty, and to provide essential connectivity to schools,

hospitals, and critical services, fostering a more equitable and informed future for humanity. We support essential

applications such as education in rural and underserved regions, telemedicine for hard-to-reach patients, seamless

connectivity for aviation and maritime users, and resilient communications during natural disasters. For example,

during the 2023 Maui wildfires, which devastated Lahaina and left thousands without power or cellular service,

138

Starlink rapidly deployed over 650 terminals to restore high-speed internet connectivity, enabling first responders,

humanitarian organizations, and survivors to coordinate relief efforts, access aid resources, communicate with

family, and support recovery in areas where traditional infrastructure had completely failed. During Hurricanes

Helene and Milton in 2024 in the southeastern United States, our Starlink terminals provided a rapid lifeline for

communication and recovery when traditional cell towers, broadband lines, and power infrastructure were knocked

out for days or weeks by widespread damage caused by flooding and high winds.

Our AI technology also has the ability to elevate the quality of life for people and communities around the world.

We believe AI has the potential to revolutionize human potential—from advanced manufacturing and infrastructure

development to scientific research and medicine—delivering tangible real-world benefits for individuals,

organizations, and governments. For example, AI systems can expedite scientific discovery for researchers, aid

healthcare professionals in precise medical analysis and diagnosis, and empower educators to craft tailored learning

experiences for students. Moreover, these technologies can optimize Earth’s resource allocation, enhance disaster

response strategies, and drive efficiencies in transportation and energy systems.

We believe that our current space efforts will catalyze transformative breakthroughs that could reshape terrestrial

industries and lead to the emergence of new trillion-dollar markets on the Moon, Mars, and beyond. In particular, we

believe our goal of establishing a lunar presence will enable terawatt-scale annual AI compute growth, support

deeper space exploration and industrialization, and serve as a stepping stone to establishing a civilization on Mars.

Due to technological advancements that we are working towards, such as in-space propellant transfer, we believe

our Starship vehicle will be capable of landing massive amounts of cargo on the Moon. Once there, we believe it

will be possible to establish a permanent presence for scientific and manufacturing pursuits. For example, we believe

that factories on the Moon will be able to take advantage of lunar resources to manufacture millions of AI compute

satellites and deploy them farther into space. Our goal is to establish a sustainable lunar presence for scientific

exploration, industrialization, and as a stepping stone to Mars, serving as a proving ground for habitats, resource

utilization, and Starship systems essential for long-term human survival beyond Earth.

We believe the next paradigm shift for humanity is the creation of a resilient, perpetually expanding spacefaring

civilization that drives continuous innovation across new frontiers, ultimately propelling us to Kardashev Type II

status—a civilization that harnesses the full energy output of our Sun. In the near term, we expect space-enabled

technologies to enhance life on Earth through greater global connectivity and breakthroughs forged in the harsh

environments of our solar system, leading to accelerating progress in energy and AI. As we build infrastructure in

the Earth’s orbit, and potentially on the Moon, Mars and beyond, we believe we are capable of unlocking an era of

unprecedented economic expansion, while also contributing to the safeguards of humanity’s future against

existential risk.

Who We Are

Our mission is to build the systems and technologies necessary to make life multiplanetary, to understand the true

nature of the universe, and to extend the light of consciousness to the stars. To do this, we’ve formed the most

ambitious, vertically integrated innovation engine on (and off) Earth. We are combining the most transformative and

critical technologies in human history, including reusable rockets, a fully global internet service, satellite-to-mobile

communications that enable connectivity everywhere, our real-time information, entertainment, and free speech

platform, and a truth-seeking AI system designed to accelerate scientific discovery and augment human capabilities.

These capabilities form a self-reinforcing ecosystem: launch systems deploy and maintain the satellite network,

which delivers ubiquitous connectivity and vast data flows; the platform surfaces real-time information and supports

open discourse; and AI processes data at scale to drive breakthroughs in physics, materials science, and space

exploration. Together, they create a foundation for the development of the infrastructure of the future and the

ultimate goal of establishing a self-sustaining human presence on other planets.

SpaceX designs, manufactures, launches, and operates the world’s most advanced rockets and spacecraft. We safely

and reliably transport astronauts, satellites, and other payloads on missions that benefit life on Earth. Since 2023, we

have launched more than 80% of mass to orbit for the world each year with an over 99% mission success rate. We

believe our unparalleled launch capabilities represent the foundational competitive advantage that enables all other

parts of our business. We operate a high-speed, low-latency broadband data and communications network powered

139

by approximately 9,600 Starlink broadband and mobile satellites in Low-Earth Orbit, delivering connectivity to

millions of consumer, enterprise, and government customers across 164 countries, territories, and other markets, as

of March 31, 2026. We also built one of the world’s most advanced models in under two years and are rapidly

scaling the associated AI compute infrastructure—starting on Earth with the goal of extending to space—at industry-

leading pace and cost efficiency. We believe that space represents the largest economic frontier in human history

and that AI is a transformative force for understanding the universe. Together, we believe that space and AI will

enable an age of abundance that will lead to an unprecedented expansion in the global economy. We are the only

company that has the foundational infrastructure across hardware and software necessary to drive transformative

innovation across space, connectivity, and AI. Our technological advancements are redefining industries on Earth,

while aiming to create new ones on the Moon, Mars, and beyond.

Our Unparalleled Launch Capabilities

Since our founding in 2002, SpaceX has cracked the code on accessing space at scale, transforming an industry

characterized by decades of stagnation, risk aversion, and economically perverse cost structures. We design,

manufacture, launch, and refurbish reusable launch vehicles that provide cost-efficient, reliable, and high-cadence

access to space for our own purposes as well as for third-party commercial and government customers. In 2025, we

launched from four primary launch pads in the United States and successfully recovered boosters across seven

landing facilities including autonomous drone ships and catch towers based on the vehicle type and mission profile.

Our extensive vertical integration and end-to-end control over the entire value chain, from design to launch to

operations, allows us to achieve unprecedented speed and cost efficiency.

As of March 31, 2026, SpaceX had launched a total mass to orbit of approximately 7,400 metric tons with an over

99% mission success rate across our Falcon rockets. We have completed approximately 650 orbital space launches,

and over 540 of those launches were completed by a flight-proven Falcon rocket. In 2025 alone, SpaceX completed

170 missions across Falcon and Starship vehicles and 159 flight-proven booster launches with an over 99% success

rate on attempted booster recoveries. We launched over 2,200 metric tons, representing over 80% of mass to orbit

for the world in 2025. With the first successful launch of Falcon 1 in 2008, we became the first private company to

successfully launch a liquid-fueled rocket to Earth’s orbit. Just two years later, in 2010, the commercial debut of the

Falcon 9 rocket revolutionized space access by delivering unprecedented cost efficiency. For example, according to

NASA, the first version of Falcon 9 in 2010 reduced launch cost to approximately $2,700 per kilogram, which

represented a reduction of approximately 85% compared to the historical average launch cost per kilogram of

$18,500. The first version of Falcon Heavy in 2018 further reduced this cost to $1,400 per kilogram, a reduction of

approximately 92% compared to the historical average. We have also reduced our internal cost of launch through a

combination of engineering improvements, manufacturing efficiencies, and economies of scale—most notably,

through our ability to drive more frequent reuse of rockets.

In December 2015, we achieved what many deemed impossible: landing a rocket launched to space back on Earth.

By 2017, we were routinely recovering and reusing the Falcon 9 first-stage booster post-launch, delivering another

step-function drop in space access costs via groundbreaking reusability. As of March 31, 2026, our Falcon 9 rockets

have demonstrated the ability to refly a first-stage 34 times. Since 2020, our Dragon spacecraft has safely flown 78

crewmembers from 20 countries. With the future deployment of Starship, which is designed to be the world’s first

fully and rapidly reusable spacecraft, we aim to reduce the cost to reach orbit by 99% or more relative to the

historical average launch cost, establishing the most affordable and scalable path to creating new opportunities in

space, such as orbital AI compute and Mars exploration.

140

Booster Reusability Enables Increasing Launch Rates

Our principal launch vehicles and spacecraft include:

•Falcon 9. As the world’s first orbital-class rapidly reusable rocket, Falcon 9 was first launched in 2010 and has

a payload capacity to LEO of approximately 23 metric tons when fully expendable. Falcon 9 has completed

approximately 620 orbital space launches as of March 31, 2026, and an over 99% mission success rate, making

it the most active orbital launch vehicle today. In 2025 alone, we launched 165 Falcon 9 rockets, of which 157

were flight-proven booster launches, and during the three months ended March 31, 2026, we launched 40

Falcon rockets, of which 39 were flight-proven booster launches.

•Falcon Heavy. Falcon Heavy first launched in 2018 when it put a Tesla Roadster and its mannequin passenger,

known as Starman, into orbit around the Sun. With a payload capacity to LEO of approximately 64 metric tons,

Falcon Heavy is a partially reusable super heavy-lift launch vehicle designed to deliver large payloads to orbit.

Falcon Heavy is one of the most powerful operational rockets in the world measured by liftoff thrust, with 11

launches as of March 31, 2026 and a 100% mission success rate.

•Dragon. Launched by Falcon 9 in 2012, our Dragon spacecraft became the first commercial spacecraft to

deliver cargo to and from the International Space Station and, eight years later, the first privately built vehicle to

fly humans to the orbiting laboratory. Since its first flight, Dragon has visited the International Space Station

over 50 times, and restored America’s ability to launch astronauts. Dragon has also supported all of NASA’s

private astronaut missions to the International Space Station, flown the first all-commercial astronaut crew,

completed the first human spaceflight over the Earth’s polar regions, and supported the first-ever commercial

spacewalk.

•Starship. First launched in 2023, Starship is designed to be a fully reusable, super heavy-lift launch vehicle.

Starship V3 is designed to deliver 100 metric tons to Earth’s orbit in a fully reusable configuration while

enabling rapid turnaround times akin to commercial aviation. Future generations of Starship are being designed

to double this payload capacity. To date, we have executed 11 Starship flight tests. We have also scheduled a

12th flight test, which will debut the next generation Starship vehicle and Super Heavy booster, powered by the

next evolution of our Raptor engine and launching from a newly designed pad at Starbase. We expect Starship

to commence payload delivery to orbit in the second half of 2026. We have achieved innovative milestones

141

such as catching a booster using “chopstick” arms on the same tower it launched from. We expect this

capability will facilitate rapid refurbishment and reuse, allowing for multiple launches per day at reduced costs.

Upon achieving rocket reusability, we recognized the immense potential of our launch business to enable new

revenue streams, as our launch capacity would eventually outstrip demand from traditional space customers alone.

This realization, along with our efforts to make life multiplanetary, drove us to reimagine what was possible when

access to space became more affordable. Rather than asking what was being done in space, we asked what large-

scale global need could be better served from space. This led to the development of Starlink, our global satellite

internet constellation, consisting of thousands of LEO satellites designed to provide high-speed, low-latency

broadband connectivity to underserved areas worldwide. Although the concept of using satellites for global internet

connectivity dates back decades, technical challenges and the prohibitive cost of accessing space historically

rendered attempts to provide such connectivity economically unviable. Within three years of our first satellite launch

in 2019, we solved the technical and production challenges of the satellites, and within five years, we had deployed

the largest LEO constellation in existence. Today, Starlink is the sole low-latency network available globally.

As the leader in space access, our launch operations are an important and expanding competitive advantage. By

combining increasing launch cadence, expanding cargo capacity, and declining unit costs—driven by rapid

reusability—we have generated a compounding competitive advantage. This not only fortifies our core business, but

also provides vast new market opportunities uniquely enabled by space.

Our Leading Capabilities Across Space, Connectivity, and AI

Space. While our launch capabilities support our other businesses, such as Starlink Consumer Broadband and

Starlink Mobile, we also sell launches to third-party customers. We offer launch services to commercial, civil, and

government customers through our reusable Falcon 9 and Falcon Heavy rockets for satellite, cargo, and crew

missions. We fly to LEO, MEO, GEO, lunar, and interplanetary trajectories, as well as the International Space

Station. We are the primary launch provider for the U.S. government. In 2025, we launched 11 of 12 National

Security Space Launch (“NSSL”) medium and heavy lift missions and all five U.S. crew and cargo missions to the

International Space Station for NASA. We serve commercial and government customers—including NASA, the

National Reconnaissance Office (“NRO”), Axiom Space, SES, Eutelsat, and Oneweb. We charge our customers

based on the type of rocket, mass to orbit, size of payload, and type of service, such as whether the launch is

dedicated to a single customer or part of a “rideshare” with other customers.

Starship is our next-generation reusable rocket vehicle that we expect will expand our launch capability dramatically

through full and rapid reusability combined with currently unprecedented mass to orbit capability. As the most

powerful launch system ever developed, we expect that Starship V3 will be able to carry a payload of 100 metric

tons, and that future generations could reach 200 metric tons, potentially as soon as Starship V4. Starship is designed

to deliver our next-generation satellites to orbit, long-haul point-to-point transportation on Earth, the cargo and crew

necessary to develop a base on the Moon and a city on Mars for research and human spaceflight development.

Connectivity. Starlink provides global access to high-speed internet, including underserved rural and remote

communities worldwide. As of March 31, 2026, we had approximately 9,600 Starlink broadband and mobile

satellites in Low-Earth Orbit, providing broadband connectivity to approximately 10.3 million Starlink Subscribers

across 164 countries, territories, and other markets. We also provide satellite-to-mobile texting and over-the-top

voice services to approximately 7.4 million monthly unique devices across approximately 30 countries.

•Starlink Consumer Broadband.We operate the world’s largest and most advanced space-based internet

broadband service with median latency at approximately 25 milliseconds as of March 31, 2026. We provide

fiber-like download speeds—at a median of 225 Mbps during peak hours for residential users as of March 31,

2026—and the technological capability to provide service everywhere on Earth, including the poles. This

service quality is enabled by our vast network of approximately 9,600 Starlink broadband and mobile satellites

in Low-Earth Orbit, which accounted for approximately 75% of all active maneuverable satellites in orbit as of

March 31, 2026. We expect to commence deploying our next-generation V3 satellites, designed to offer one

Tbps of downlink capacity per satellite, using Starship in the second half of 2026. We expect that a single

Starship launch will be capable of deploying up to 60 V3 satellites to LEO, representing a potential twenty-fold

142

increase in Starlink downlink capacity deployed relative to a Falcon 9 launch. As of March 31, 2026, we had

approximately 10.3 million Starlink Subscribers, up approximately 105% from 5.0 million subscribers a year

prior. We charge our Starlink Subscribers a monthly subscription fee, which varies based on geographic market

and download speed, plus typically a one-time upfront terminal cost.

•Enterprise Solutions. SpaceX is a critical partner to a wide array of enterprises. We offer Starlink’s high-

speed, low-latency, reliable internet services to enterprise customers across industries including construction,

agriculture, retail, telecom, hospitality, aviation, maritime, and land mobility. Starlink’s unique capabilities are

well‑suited for deployments across field offices, remote worksites, research stations, drilling rigs, rural

hospitals, aircraft, cruise ships, trains, and hotels. Our enterprise customers include companies such as United

Airlines, Carnival, Maersk, and John Deere, among others. We also serve a broad fixed‑site customer base

across industries such as retail and financial services that require high availability for critical operations as well

as reliable connectivity in remote or hard-to-serve locations. As companies continue to invest in secure and

resilient networks and backup systems to keep critical infrastructure online—such as point‑of‑sale and payment

processing systems—we often start as a backup solution and then transition to being the primary solution. Our

enterprise contracts are based on a combination of subscriptions, data consumption, capacity, or other pricing

models depending on each customer’s particular needs. Since 2023, no Starlink Enterprise customer having

contributed more than $750,000 of annual revenue has voluntarily discontinued their service, demonstrating the

strong performance and value of our offering. This is despite the ability of our customers to cancel the service at

any time.

•Government Solutions. For our government customers, we provide high-speed, resilient connectivity for

public services, social impact, humanitarian efforts, and disaster response in even the most remote and

challenging environments. Examples include support for the FEMA in coordinating disaster recovery after

hurricanes and wildfires, the NOAA for at-sea testing and environmental monitoring, the Government of the

Philippines for linking remote islands, schools, and public institutions, the Government of Jamaica for

improving digital access in remote and maritime areas, and the Government of Ecuador for supporting

education and healthcare connectivity in isolated communities. Separately with Starshield, we have leveraged

our commercial LEO satellite constellation engineering learnings and operational experiences to develop a

secure, dedicated satellite network designed specifically for United States Government customers and national

security applications.

•Starlink Mobile. We provide satellite-to-mobile connectivity, supplementing terrestrial networks and

substantially reducing mobile “dead zones” across approximately 30 countries. We partner with MNOs

including major wireless carriers like T-Mobile in the United States, and other international operators including

One NZ, Optus, Telstra, Rogers, KDDI, Salt, Entel, Kyivstar, and VMO2. Through these partnerships, we

enable consumers, businesses, and public-sector customers to use their existing phones in more places, support

critical connectivity during disasters and power outages, and open new applications for low-bandwidth mobile

and IoT devices. Our current capabilities under our “V1” constellation (consisting of approximately 650 V1

Mobile satellites in orbit) include light data, text messaging (SMS), and over-the-top voice services (e.g.,

WhatsApp and FaceTime). We are developing more comprehensive satellite-to-mobile services, including

broadband data and IoT connectivity, which are expected to deliver resilient, infrastructure-independent

connectivity worldwide and enable 5G connectivity. We have partnerships with approximately 30 MNOs on six

continents, covering an area that is home to approximately 1.9 billion people. We charge MNOs either a fixed

fee or a per-mobile user fee-based amount, which is typically passed through to the customer via the carrier as

an “add-on” feature.

143

Our Global Starlink Subscriber Base

AI. We operate a highly vertically integrated AI platform spanning gigawatt-scale AI compute infrastructure, our

truth-seeking frontier AI model, Grok, AI solutions for consumer and enterprise customers, and X, our real-time

information, entertainment, and free speech platform. We believe AI is rapidly converging toward AGI, where

human cognitive capabilities can be replicated and scaled at machine speeds, profoundly augmenting human

productivity. Once an AGI system exists, its true value derives from the ability to create limitless duplicates of

human-like intelligence, necessitating vast computational resources and cost-efficient deployment to achieve

meaningful scale. Without large-scale, power-efficient infrastructure, AGI cannot be deployed broadly or

economically—making such infrastructure a critical strategic differentiator.

•AI Compute Infrastructure. xAI has established a leading position in building and scaling terrestrial AI

compute infrastructure, becoming the first company to deploy a coherent gigawatt-scale AI training cluster. Our

AI compute facilities, COLOSSUS and COLOSSUS II, collectively provide approximately 1.0 gigawatt of

compute power, with additional power capacity available for data center operations. Our first-principles

thinking enables us to build coherent compute at scale and at rapid speed with lower costs than most other

companies in the industry. We brought the first cluster of COLOSSUS online in 122 days, repurposing the shell

of an existing factory, and the first cluster of COLOSSUS II online even faster in 91 days. As an illustrative

comparison, an industry benchmark to bring online a 100 megawatt greenfield data center is approximately two

years. We also demonstrated a significant improvement in cost efficiency, achieving data center construction

costs for COLOSSUS II that are considerably lower than industry benchmarks on a per megawatt basis. This

dual speed and cost advantage stems from our complete vertical integration and the shared culture infused by

our founder, Mr. Musk, across our Space, Connectivity, and AI segments. The addition of Terafab, an initiative

together with Tesla to build a manufacturing facility capable of producing 1 terawatt per year of compute

hardware, aims to further extend our vertical integration to chip design and manufacturing to alleviate potential

future chip shortages at SpaceX, optimize compute performance, and potentially reduce overall compute costs.

Intel, which has the ability to design, fabricate, and package ultra-high-performance chips at scale, has also

joined the Terafab project. We believe that the key constraints in the continued growth of AI are physical—chip

manufacturing, data center infrastructure, and power generation; the future of AI will be determined by the

control of the physical stack.

144

•Truth-Seeking Frontier Model. xAI has developed one of the world’s most advanced, truth-seeking frontier

models with Grok. Since launching Grok-1 in November 2023, we have released four major versions and

notable variations thereof, achieving one of the fastest iteration cycles in the industry, culminating in Grok-4.3

(April 2026). Building on this trajectory, we expect to continue scaling Grok through subsequent generations.

Ongoing training of next‑generation models is expected to scale toward multiple trillions of parameters, which

could represent a step change in reasoning in depth and overall intelligence. In this context, the number of

parameters refers to the scale of the model, where parameters are the internal numerical values, such as

“weights,” that are adjusted during training to enable the model to recognize patterns and relationships in data.

A larger number of parameters generally allows the model to capture more complex relationships, store greater

amounts of knowledge, and achieve higher levels of reasoning capability.Within two years of its initial model

release, Grok achieved frontier-level performance in scientific reasoning, as measured by its GPQA Diamond

score, an industry benchmark that evaluates AI models on a standardized set of questions written and validated

by experts, on a faster timeline than reported by other leading model providers. This accelerated rate of

innovation stems from our highly vertically integrated stack: full ownership of training infrastructure, access to

the world’s most powerful compute clusters, and relentless focus on truth seeking and real-world utility. A key

competitive differentiator is Grok’s deep integration with X, enabling proprietary access to a real-time

information stream of approximately 350 million daily posts, which enhances freshness, relevance, and

contextual awareness for Grok. This direct, real-time access to the information and human discourse on X

enhances Grok’s truth-seeking capabilities by grounding outputs in up-to-date knowledge and diverse

viewpoints. We believe that this combination of compute infrastructure scale and the massive dataset available

to us through X, subject to some limitations for certain content, has allowed us to achieve industry-leading

performance and provide model outputs that analyze real-time information on global events. We expect that our

compute infrastructure and direct access to real-time data via X constitute substantial performance advantages

for Grok that will result in increasingly rapid and dramatic iteration cycles.

•Consumer and Enterprise Applications. We leverage our leading frontier models and compute infrastructure

to deliver consumer and enterprise applications. In under six months, we developed Grok Voice, a real-time

speech engine, including in multilingual performance. Our image and video generation system, Imagine,

produced approximately 10 billion images and over 2 billion videos per month, on average, for the quarter

ending March 31, 2026. Together with Tesla, we are also developing Macrohard, an agentic AI platform

designed to be capable of fully emulating digital workflows and augmenting human operation of computers—

from coding and product development to management and entire business processes—using sophisticated

autonomous agents. We believe Macrohard will have the potential to fundamentally transform how companies

are structured and operate, thereby allowing dramatic increases in human productivity. In addition, we believe

our existing government relationships and track record as large government contractors are a structural

advantage as governments become significant consumers of AI applications.

Our integrated AI platforms across Grok and X have over 1.3 billion supported accounts active in the last

twelve months ended March 31, 2026, including approximately 550 million MAUs, up from over 1.1 billion

supported accounts and approximately 520 million MAUs as of December 31, 2025. Of our MAUs, we had

approximately 117 million MAUs that used Grok’s AI features as of March 31, 2026.

We also monetize user activity through high-impact advertising inventory on X. We believe X’s scale, real-time

engagement, and integration with Grok provide a differentiated foundation for building a unified user

experience across communication, content discovery, commerce, and financial services, among others. For

enterprises that advertise on X, we offer large-scale user engagement, real-time content, and advanced AI-

driven performance marketing tools. For enterprises, we offer tailored deployments of Grok customized to

specific workflows and security needs through Grok Business and Grok Enterprise, sold on license-,

consumption-, or outcome-based pricing models.

Collaboration with Tesla

SpaceX and Tesla developed the early foundation of a strong and constructive partnership through a series of limited

but successful commercial engagements. Our relationship with Tesla evolved meaningfully following Tesla’s

January 2026 commitment to invest in xAI—an investment that, upon SpaceX’s acquisition of xAI, was converted

145

into an equity interest in SpaceX. Tesla and xAI continue to build upon their longstanding collaborative relationship

by evaluating future strategic opportunities between the companies.

One expected area of collaboration is an AI project called Macrohard. We expect Macrohard to benefit from running

on both state-of-the-art processors and cost efficient, next-generation Tesla processors, a critical advantage of our

vertical integration.

Another expected area of collaboration is Terafab, an announced AI chip manufacturing initiative designed to

vertically integrate the design, fabrication, and deployment of advanced logic and memory chips. We believe this

initiative will alleviate potential future chip shortages at SpaceX and optimize compute performance.  We expect

Terafab to be the world’s largest chip manufacturing facility. Our strategy for Terafab is to vertically integrate

across the design of lithography masks, fabrication of logic and memory chips, and design of advanced packaging in

a single closed-loop plant. Conducting all these activities end-to-end in a single facility enables rapid testing and

iterations, allowing us to improve chip design and scale manufacturing faster. We expect that our speed and cost

advantage from vertical integration will allow us to scale efficiently in AI chip manufacturing towards our long-term

goal of producing one terawatt of compute each year. We are partnering to build Terafab in order to support growth

in two kinds of chips— one type optimized for terrestrial edge and inference to be used primarily in Tesla’s Optimus

robots and vehicles, and another type optimized for the space environment to be used in our orbital compute

infrastructure. While Terafab is intended to expand our internal chip manufacturing capabilities, we expect to

continue sourcing a significant portion of our compute hardware from third-party suppliers. We view Terafab as

complementary to these relationships, enabling us to augment our access to compute hardware at massive scale and

further complete our highly vertically integrated compute platform by extending our control to the foundational chip

layer. We believe that the key constraints in the continued growth of AI are physical—chip manufacturing, data

center infrastructure, and power generation; the future of AI will be determined by the control of the physical stack.

We believe that we are better positioned than other AI companies given our unique control over the full physical

stack. We plan to explore other areas of strategic collaboration with Tesla in the future.

Collaboration with Cursor

On April 19, 2026, we entered into a compute agreement with Cursor. Cursor develops and operates an AI-native

integrated development environment that enables professional software developers and engineering teams to write,

edit, review, and refactor code using LLM-powered agents and workflows integrated via its proprietary model

harness. In 2025, Cursor launched Composer, its own LLM trained for software development. It recently released

Composer 2, which offers improvements in coding performance at lower cost. We believe the compute agreement

and any acquisition of Cursor (described below), if completed, will extend our strategy to vertically integrate

compute infrastructure, models, and applications, can help accelerate our development of AI-native software tools,

and combined with our significant compute capacity, will help strengthen our position in AI-assisted developer

productivity. We expect to accelerate the development of our existing AI models, including Grok, through our

collaboration with Cursor.

Under the compute agreement, we will provide Cursor with certain GPU cluster compute capacity for use in

connection with specified development, training, improvement and other activities related to AI models and other

technology and intellectual property. In exchange, Cursor will contribute certain personnel, data and datasets,

documentation, technical know-how, workflows, prompts, specifications and software code. We will collaborate

with Cursor to improve our existing models, including Grok, and potentially to jointly develop AI models and

related model-specific deliverables. Each party retains ownership of its pre-existing and independently developed

intellectual property (including, in the case of SpaceX, Grok) and related improvements and derivatives, including

where they are utilized in connection with joint development activities. Any jointly developed models will be jointly

owned, and each party will have a broad right to use, reproduce, modify, distribute, license, commercialize and

otherwise exploit them without an obligation to account to the other party.

We also entered into an option agreement pursuant to which we have the right, but not the obligation, to acquire

Cursor. The option agreement generally provides that we may exercise the call option at any time during the 30-day

period following the earlier of (i) seven trading days following the completion of this offering and (ii) September 30,

146

2026. Exercise of the call option is in our sole discretion and subject to further approval by our board of directors.

Cursor is also subject to certain exclusivity obligations under the option agreement.

If we exercise the call option, we would simultaneously execute a merger agreement with Cursor, pursuant to which,

following satisfaction of the closing conditions set forth in the merger agreement, including receipt of requisite

regulatory approvals, Cursor would become our subsidiary, and, as a result, we would acquire all of Cursor’s cash,

intellectual property, personnel, customer contracts and other assets. As of January 31, 2026 (Cursor’s fiscal year-

end), Cursor had $3.1 billion of total assets, primarily comprising $2.7 billion of cash and cash equivalents, and

$0.55 billion of total liabilities. The purchase price would primarily be allocated to goodwill on our balance sheet.

Cursor has historically earned some revenue by providing services to customers and, if we acquired Cursor, we may

provide these or similar services to customers after the acquisition although at revenue levels that may vary

significantly from historical performance. If we exercise the call option to acquire Cursor, we would expect to retain

certain Cursor talent by committing to provide continuing employees with competitive compensation and retention-

focused incentives designed to support the long-term value of SpaceX.

The consideration for the acquisition of Cursor, if any, after the closing of this offering would consist of shares of

our Class A common stock based on an implied equity value of Cursor of $60.0 billion, and the price of our Class A

common stock that equals the volume-weighted average closing price thereof over the seven consecutive trading

days immediately preceding the closing of the acquisition. If either (i) we decide to terminate the option agreement

or (ii) Cursor is eligible to and decides to terminate due to our material breach of the option agreement (subject to

notice and cure provisions), Cursor is entitled to a $1.5 billion termination fee under the option agreement and an

$8.5 billion deferred services fee under the compute agreement. These fees are payable in cash (or Class A common

stock, if this offering has not been consummated at the time the fees become payable).

Any shares of our Class A common stock issuable pursuant to the merger agreement would be issued in reliance

upon the exemption from the registration requirements of the Securities Act provided by Section 4(a)(2) thereof. As

a result, any such shares of Class A common stock would be deemed “restricted securities” as such term is defined

under Rule 144 under the Securities Act. Such shares of Class A common stock would be eligible for resale only if

registered under the Securities Act or if such resales qualify for an exemption from registration.

We have conducted preliminary due diligence on Cursor’s business, technology and operations, and expect to

continue such diligence in connection with any decision to exercise the call option. We cannot predict whether we

will elect to exercise the call option or, if exercised, whether the acquisition will close on the anticipated terms, or at

all.

Compute Services Agreements with Third Parties

We believe our compute infrastructure and related strategy provides us with substantial flexibility in how we

allocate and monetize capacity. We have the ability to use compute resources to support our proprietary AI

applications (such as Grok 5, which is currently being trained at COLOSSUS II), while also providing access to

select compute capacity to third-party customers. For example, in May 2026, we entered into Cloud Services

Agreements with Anthropic, an AI research and development public benefit corporation, with respect to access to

compute capacity across COLOSSUS and COLOSSUS II. Pursuant to these agreements, the customer has agreed to

pay us $1.25 billion per month through May 2029, with capacity ramping in May and June 2026 at a reduced fee.

The agreements may be terminated by either party upon 90 days’ notice. The customer will retain ownership and

intellectual property rights in its content, AI models, and related data. This structure allows us to monetize unused

compute capacity in our infrastructure, while still permitting reallocation of the capacity for our own internal

initiatives if needed in the future. We have sufficient capacity to provide compute for our own AI models, including

support of our training and inference demands, and to satisfy the obligations under these agreements. We expect to

enter into additional similar services contracts for compute capacity with third parties. To the extent we become

compute constrained due internal and external utilization, we would need to expand our compute infrastructure. We

believe this opportunity highlights the increasing importance of large-scale, frontier-level AI infrastructure and

positions us as a differentiated provider of high-performance compute capacity to both internal and third-party AI

workloads. We believe our dual monetization strategy provides multiple pathways to generate returns on invested

capital.

147

Our Repeatable Business Model

Our business model is built on a repeatable, engineering-driven framework that combines our unparalleled launch

capabilities, extreme vertical integration, rapid iteration, and disciplined capital investment to create durable, large-

scale businesses. We execute this framework through the following core principles:

1.Leverage our unparalleled launch capabilities to enable massive scale. Our rockets—with unmatched

launch cadence, best-in-class reliability, and dramatically reduced cost-to-orbit—are the foundation that we

expect will enable us to create economic opportunities in space and deliver a diversified portfolio of services.

Our launch capabilities enable large-scale deployment of assets that would not otherwise be economically

viable.

2.Identify and create new trillion-dollar market opportunities. We focus on market opportunities that are

useful for humanity and that present trillion-dollar opportunities, including global broadband and mobile

connectivity for consumers, enterprises, and governments; and AI applications and computational infrastructure.

We prioritize opportunities where structural inefficiencies or legacy technological limitations have constrained

supply.

3.Design a solution with world-class engineering and first-principles thinking. We apply physics-based

engineering and first-principles thinking to design products and systems from the ground up—boiling things

down to the most fundamental truths and reasoning up from there. This helps us drive massive, step-function

improvements in performance, scalability, and cost.

4.Apply “The Algorithm” (make less dumb, delete, optimize, accelerate, automate). We operate under a set

of core execution principles that we refer to as “The Algorithm,” a five-step iterative process that we use as our

guiding principles day-to-day. We make the requirements less dumb, delete unnecessary processes or parts

(embracing the principle that the best part is no part), only then optimize the necessary processes or parts, and

then accelerate cycle time (many entities have launched once; no one other than us has ever launched over 100

times per year), and automate only proven processes after the first four steps are completed. We apply the

Algorithm across every aspect of our organization, creating a cultural and operational standard of excellence

that has defined SpaceX since inception.

5.Vertically integrate all the way to the end customer.We design and manufacture a significant portion of our

components in-house, including engines, avionics, structures, and software, even producing the “tools that make

the tools,” enabling us to test, fail, and iterate rapidly. We can then release newer, more advanced hardware with

speed and cost efficiency.

6.Continuously drive cost down and throughput up. Through rocket reusability, manufacturing at scale,

advanced automation, and rigorous operational discipline, we continuously reduce unit costs while increasing

launch cadence, satellite network, and AI hosting capacity.

7.Generate significant cash flow and reinvest in the future. As our businesses scale, they generate significant

cash flow, which we reinvest into nascent market opportunities—driving a self-reinforcing cycle of constant

innovation and potentially creating significant additional value.

Starship is a powerful example of this business model in action. Upon achieving a fully and rapidly reusable design,

we believe Starship will support a step-function increase in launch capacity and be capable of landing massive

amounts of cargo on the Moon. Once there, we believe it will be possible to establish a permanent presence for

scientific and manufacturing pursuits. For example, we believe that factories on the Moon could take advantage of

lunar resources to manufacture millions of AI compute satellites and deploy them farther into space. Additionally,

we are collaborating with NASA under the Artemis program to land humans on the Moon, with the goal of using

Starship for transportation, which will be the first such mission since 1972.

We will continue leveraging our expanding launch capabilities, combined with our engineering and manufacturing

expertise, to create and scale new markets in space for the benefit of humanity—on Earth, the Moon, Mars, and

beyond.

148

Our Engineering-First Culture

We are able to achieve transformative technological breakthroughs because we accept only the laws of physics as

the limiting factors to our work and mission. Our core approach is deeply rooted in first-principles thinking, which

rejects any preconceived notions or experience-based norms. Our unparalleled track record demonstrates our

capacity to execute space missions and achieve technological breakthroughs with speed and precision that others

have not achieved. We have a track record of achieving what many have deemed impossible. Some of our industry-

defining achievements and historic milestones include:

•The first private company to develop and launch a liquid-fuel rocket to reach orbit (2008);

•The first private company to successfully dock a private spacecraft with the International Space Station (2012);

•The first to successfully propulsively land (2015) and refly orbital-class rocket boosters (2017);

•The first to begin deploying a large-scale LEO broadband satellite constellation (2019);

•The first private company to transport astronauts to orbit, returning America’s ability to fly astronauts to and

from the International Space Station (2020);

•The first to manufacture consumer-grade phased-array user terminals at scale (2022);

•The first to deploy a large-scale LEO satellite-to-mobile constellation (2025);

•The first to build a gigawatt-scale AI training cluster and largest coherent supercomputer (2026);

•The first gigawatt-scale Megapack battery installation (2026); and

•The only company capable of building orbital AI compute at scale.

Our organizational philosophy fosters an engineering- and data-led culture that embraces failure as an essential

learning opportunity and is maniacally focused on efficiency and speed. This culture allows us to deliberately move

quickly to test new hardware, knowing that early failures provide more valuable data than protracted analysis. We

view our factories as the machines that build the machines and maintain a relentless focus on our ability to move,

fail, and fix fast.

Our AI Compute Infrastructure Advantage and Growth Strategy

We believe AI leadership will be defined by the ability to rapidly scale compute capacity to support exponential

usage growth and frontier intelligence. There is a meaningful compounding benefit of greater usage, creating more

data for training, driving improvements in model performance, and in turn leading to greater usage. We believe that

our highly vertically integrated, shovels-to-tokens approach allows us to train and iterate our frontier models at

lower cost and higher velocity, accelerating development cycles, eliminating external bottlenecks, and driving rapid,

continuous improvements in model performance. This dynamic reinforces the criticality of scale and cost efficiency

in compute infrastructure as the primary differentiator in the AI landscape. In addition, our leadership in compute

infrastructure positions us to monetize not only AI software applications built on our models, but also the underlying

compute that powers them. As we continue to scale our terrestrial and orbital compute infrastructure to support

internal model development, training, and inference workloads, we intend to sell our high-performance compute

capacity to a limited number of third party customers.

Why Compute Matters. The training and inference demanded by advanced AI models require substantial

computational resources. Greater compute capacity enables more intelligence by training new generations of models

with increasing frequency and creating more capable models, ability to support inference, or usage, across a large

and growing user base, and extraction of the highest performance from those models. As the AI user base expands,

we also expect compute demand per user to increase significantly. Reasoning models introduced in 2024

demonstrated that allocating more computational resources during inference directly leads to higher-quality

intelligence. AI agents popularized in 2026 demonstrated that allocating more computational resources enabled

149

multi-step task execution, meaningfully increasing compute demand per human user interaction. In addition,

compute infrastructure with end-to-end, cluster-level coherence through tight integration across software and

hardware systems enables more efficient, stable, and higher-fidelity training and inference at scale—ultimately

enhancing model intelligence and performance. Within inference, we expect computationally-intensive reasoning,

agentic, and multi-modal workloads will continue to grow as a portion of overall usage. We therefore expect

demand for compute will continue to increase across consumer, enterprise, and government applications as AI

adoption accelerates. For example, U.S. compute demand has already outpaced available power supply with

estimated demand of 62 gigawatts in 2025 exceeding the power generation of 49 gigawatts, according to industry

sources. We expect the gap between demand for compute and power supply to continue to widen meaningfully as AI

compute needs proliferate.  Furthermore, we believe that third-party estimates on data center demand are constrained

by the practical supply limitations that exist in a terrestrial context and the power shortage may be far greater than

what research estimates suggest. We believe operators with superior model-to-compute integration—the ability to

efficiently support and allocate compute across both training and inference workloads—are best positioned to win

the AI race.

Self-Reinforcing Network Effects Among Lower Cost Per Token, Model Quality, and User Adoption. AI systems

are ultimately constrained or differentiated by the cost, speed, and scale at which they can generate and process

tokens. A “token” represents the fundamental unit of data consumed and produced by modern AI models, for

example corresponding to words, images, audio, or other modalities. It serves as the atomic unit through which

models read, reason, and generate output. As such, tokens are the primary basis for measuring both the cost of

training and cost of inference, making them a foundational economic metric in the AI space. Companies that can

structurally reduce energy, compute, networking, and deployment costs per token will be positioned to train faster,

iterate more rapidly, and ultimately manufacture greater intelligence, scale models more rapidly, and deliver

increasingly powerful and accessible AI solutions. This creates a self-reinforcing advantage in which lower token

costs drive greater model quality and user adoption, reinforcing AI leadership. This is because lower cost per token

enables more frequent model training, larger and more sophisticated models, longer chains of processing for

reasoning and agentic workloads, and significantly higher inference volumes at economically viable prices. This

dynamic directly impacts model quality, responsiveness, and accessibility, while also determining the ability to

serve the rising global demand across consumer, enterprise, and mission-critical AI applications. As AI systems

scale toward increasingly complex reasoning tasks and higher usage intensity, improvement in cost per token

enables meaningful advantages in performance quality, scaled distribution, and monetization. This is particularly

true as the industry converges towards recursive self-improving learning that minimizes human intervention, which

is highly token consumptive.

Cost of Compute is the Main Driver of Cost Per Token. The cost of compute is the primary driver of cost per token

across both training and inference workloads. Each token processed by an AI model requires a quantifiable amount

of computational effort. The total cost per token is determined by the efficiency, availability, and unit economics of

the underlying compute resources. According to SemiAnalysis, for most AI companies without a build cost

advantage, their total capital cost of building compute infrastructure derives approximately 30% from data center

construction costs (including, but not limited to, the shell; mechanical, electrical, and plumbing (“MEP”); and grid

interconnection) and approximately 70% from the cost of procuring processors and critical IT equipment. Ongoing

operational costs of utilizing this compute infrastructure include the cost of power to run the processors, cost of

maintaining those processors, and cost of delivering inference workloads to the end user. Improvement in the cost of

building and operating this compute infrastructure—whether through lower data center construction cost, lower

power infrastructure cost, shorter time to grid interconnection, or higher cluster-level throughput—translates directly

into lower cost per token. Accordingly, for a given level of intelligence, we expect the long-term economics of AI

companies to be driven by the ability to consistently deliver bleeding-edge compute at the lowest possible cost per

token. Put simply, we view cost per token as a function of three primary inputs—the underlying AI model, the

compute hardware, and energy, and we expect to have a competitive advantage in the latter two cost components.

We believe we have a pathway over time that will significantly reduce compute hardware costs through continued

vertical integration and development of proprietary chips, building on our experience designing custom silicon for

our Starlink satellites. We also expect that the marginal cost of energy for our AI compute satellites will be minimal

because our satellites are powered by solar arrays in space. By driving the energy component to minimal levels and

150

pursuing improvements in compute hardware cost, we believe we can achieve a meaningfully lower overall cost per

token in the future.

We Have a Dual Speed and Cost Advantage in Terrestrial AI Compute. We have established a leading position in

building and scaling terrestrial AI compute infrastructure, becoming the first company to deploy a coherent

gigawatt-scale AI training cluster. We own and operate what we believe to be the largest AI training data center

clusters on Earth. Our AI compute facilities, COLOSSUS and COLOSSUS II, collectively provide approximately

1.0 gigawatt of compute power, with additional power capacity available for data center operations. Our first-

principles thinking enables us to build coherent compute at scale and at rapid speed with lower costs than most other

companies in the industry. We brought the first cluster of COLOSSUS online in 122 days, repurposing the shell of

an existing factory, and the first cluster of COLOSSUS II online even faster in 91 days. As an illustrative

comparison, an industry benchmark to bring online a 100 megawatt greenfield data center is approximately two

years. We also demonstrated a significant improvement in cost efficiency, achieving data center construction costs

for COLOSSUS II that are considerably lower than industry benchmarks on a per megawatt basis. We are able to

deploy power and compute significantly faster than other AI companies through first-principles thinking, behind-

the-meter power generation, coupled with what we believe is the world’s largest network of sustainable battery

storage systems, and innovations in advanced liquid cooling, high-density rack layouts, and efficient networking.

Our first-principles thinking and innovations in advanced liquid cooling, high-density rack layouts, and efficient

networking enable rapid, cost-effective scaling with the latest processors—keeping us ahead of competitors

deploying traditional methods. Faster deployments reinforce our cost advantage: we are able to access and bring

online the highest performing hardware before our competitors, allowing us to sustain a token cost advantage. For

example, we believe COLOSSUS II became one of the world’s first data centers to deploy GB200s and GB300s at

significant scale and is currently powering training for our next frontier models, including Grok-5. We have already

proven in multiple large-scale terrestrial data centers that we have built not only faster than competitors in the

industry, but also at a lower cost.

We have a Unique Right to Win in Orbital AI. The Sun contains approximately 99.8% of the solar system’s energy

and offers what we believe is the only truly scalable solution to terrestrial energy constraints, as we expect the cost

and availability of terrestrial energy sources over time will necessitate a transition to orbital AI solutions. The logical

path forward is to move power-intensive AI workloads into orbit, where solar energy is near-constant and

uninterrupted. With such accessibility to energy, we believe that our launch business will enable us to consistently

activate the highest performing hardware before our competitors without such access, shrinking the timeline to

useful tokens on bleeding-edge hardware and sustaining our token cost advantage. Manufacturing next-generation

satellites and launching them into space in very large numbers is a core component of our plans. We believe we are

the only company with a commercially viable path to building orbital AI compute at scale. This is underpinned by

our unique ability to launch substantial mass into orbit cost-efficiently through reusable rockets and to manufacture

secure, reliable, and high-performance satellites at low cost and high volume.

•Terrestrial compute leadership. We believe the same cost and build advantages that have underpinned our

leadership in gigawatt-scale terrestrial data centers will enable us to innovate across other terrestrial data center

formats such as modular data centers for inference. We believe our modular terrestrial data center architectures

will provide a foundation for the deployment of compute infrastructure in orbit given similarities in form factor

in contrast to a gigawatt-scale campus.

•Satellites. Just as we expect our expertise in terrestrial data centers will enable us to package AI compute into

modular, satellite form factors, we expect our leadership in satellite communications to allow us to interconnect

our fleet of AI compute satellites into a massive, coherent constellation of compute. For example, as of March

31, 2026, our constellation already incorporated over 23,000 inter-satellite lasers that create a dynamic mesh

network in space, enabling traffic to route through orbit rather than relying solely on terrestrial backhaul

infrastructure. We are designing next-generation, high-performance AI compute satellites built for high volume,

low cost, and with the reliability required for long-duration operation in space.

•Starship. We expect each of our Starship V3 vehicles to carry 100 metric tons to Earth’s orbit in a reusable

configuration, and future generations could reach 200 metric tons in capacity, potentially as soon as Starship

151

V4. Future generations of Starship are being designed to eventually deliver millions of tons to orbit and beyond

per year. Delivering large amounts of mass to orbit at low cost will be critical to deploying AI compute satellites

at scale.

We Believe Orbital AI Can Accelerate Time to Power and Reduce Token Costs. The Sun contains approximately

99.8% of the solar system’s energy and offers what we believe is the only truly scalable solution to the challenge of

accelerating demand for compute relative to terrestrial energy constraints. The logical path forward is to move

power-intensive AI workloads into orbit, where solar energy is near-constant and uninterrupted. With such

accessibility to energy, we believe that our launch business will enable us to consistently activate the highest

performing hardware before our competitors without such access. We believe SpaceX is uniquely positioned to

deploy and operate data centers in orbit that can eventually achieve a lower cost than terrestrial data centers over

time due to our extreme vertically integrated approach across launch, satellite manufacturing at scale, network

connectivity and terrestrial data center expertise.

•Time to useful tokens on new generations of infrastructure.Although we have already demonstrated an

ability to rapidly scale new generations of compute in terrestrial deployments, we believe orbital AI will

accelerate our time to useful tokens on bleeding-edge AI infrastructure. Physical deployment of new hardware

is expected to be enabled by our launch business, where we believe reusability and launch cost efficiency will

drive rapid cycles of payload delivery. Rapid time to useful tokens on that hardware will be enabled by the

Sun’s near-constant, uninterrupted supply of power, which would circumvent terrestrial power infrastructure

constraints such as power procurement, grid interconnections, and permitting. As new generations of AI

infrastructure continue to deliver step-function improvements in token efficiency, we believe that maintaining

an AI fleet consistently at the bleeding edge of the frontier curve has the potential to deliver a sustainable cost

per token advantage relative to our competitors.

•Construction, power, and cooling infrastructure. In orbit, construction costs are replaced by launch costs and

satellite production costs. We expect reusable launch systems and high flight cadence will significantly reduce

the cost per kilogram to orbit, enabling more efficient deployment of compute payloads to orbit, and eventually

approach the cost of fuel. We believe our advanced satellite manufacturing capabilities enable us to build AI

compute satellites at scale and lower cost than competitors. Other terrestrial data center construction costs such

as building the shell, MEP, and grid interconnection are not applicable in space. As a result, once Starship and

our AI compute satellites are fully deployed at scale, we believe that the initial deployment costs of in-orbit

compute in the aggregate will be less than construction costs of others’ terrestrial data centers.

•Cost to procure and service processors. The cost of processors is a significant cost for both terrestrial and

orbital data centers. We do not believe that moving compute to space in and of itself will have a meaningful

impact on the cost of procuring processors. However, we believe that diversifying our long-term access to the

supply of processors, including through our Terafab initiative with Tesla and Intel, will be a key driver in

reducing the overall cost of compute hardware over time. By combining internally manufactured, lower cost

chips with those we source from third-party suppliers, we expect the overall cost of our processors to decline. In

addition to reducing costs, we also expect that this hybrid sourcing strategy will help alleviate potential future

chip shortages at SpaceX. In addition, we intend to conduct intensive pre-deployment testing to reduce the rate

of chip failure in space, as we do not anticipate servicing or repairing processors in space.

•Ongoing operations. The total cost of operating data centers is heavily influenced by energy, cooling, and

distribution requirements. In orbit, chips are expected to be powered by solar energy which is low cost and

unlimited, and we expect to leverage radiative cooling architectures, which incur no operating costs compared

to liquid or air cooling. Our integrated, space-based Starlink network architecture also enables more cost

efficient routing of data between compute clusters and to end users on a global basis.

We Believe We Are Well-Positioned to Deliver Orbital AI Compute. We believe orbital AI compute is an incredibly

difficult technical challenge that only we can solve at scale in the near term. We are the only company that has

already accomplished the key technical challenges associated with evolving connectivity satellites into AI compute

satellites. In our view, due to our proven experience, we are well-positioned to deliver a full-scale AI compute

satellite constellation. Significant work remains, but we are confident in our singular leadership position.

152

•We have unmatched satellite launch capabilities to enable deployment at scale. Our ability to launch mass

at scale and low cost is our foundational competitive advantage. Deployment of 100 gigawatts per year via

satellites carrying over 100 kilowatts of compute power per metric ton will require thousands of launches per

year and the transport of approximately one million metric tons to orbit annually. The fully reusable nature of

Starship positions us to be capable of launching this level of mass. We plan to leverage our PEZ dispenser

system, an integrated payload deployment system for Starship, along with our experience in developing fully

deployable single-unit systems that are designed to substantially reduce the risks associated with in-orbit

assembly. Starlink Broadband V1 and V2 Minisatellites have already demonstrated launch survivability and

high reliability under vibration, shock, g-loads, acoustic stress, and vacuum exposure, achieving 99.9% average

uptime. Although introducing AI processors would traditionally increase component-level failure rates, we plan

to subject compute hardware to extensive pre-deployment testing on Earth to identify early life failures before

launch.

•We have already solved many of the significant technical hurdles to evolving connectivity satellites into

AI compute satellites. Through our leading expertise in connectivity satellites and Starlink’s existing technical

and operational capabilities—including constellation-scale satellite management, autonomous operations, over-

the-air software updates, inter-satellite laser communications, mesh network deployment, radiation-hardened

system design, proprietary chip development, and the ability to operate computers reliably in the space

environment—we have already solved the hardest part in the development of AI compute satellites. AI compute

satellites represent an evolution of spacecraft engineering already demonstrated at scale through Starlink’s

connectivity satellites, and we believe development of AI compute satellites will be easier for us than for

anyone else. AI compute satellites must integrate high-density compute payloads developed with radiation-

tolerant designs and components with high electrical power generation, advanced thermal management, and

inter satellite networking. To source the electricity needed to power our AI processors, we aim to continuously

scale our existing space-grade solar technologies through insourced process development and build a

constellation in dawn-dusk Sun-synchronous orbit that delivers near-constant solar exposure. We expect solar

cells optimized for the space environment will be produced at a rapid rate, with early satellites generating 100

kilowatts of compute power and scaling from there. In orbit, thermal control must be accomplished through

radiation rather than convection and conduction. We plan to advance thermal control systems—many of which

have been proven on Starlink—by using radiators, vapor chambers, active cooling loops, and coatings to

dissipate the heat generated by AI hardware in space’s vacuum. We will also utilize inter-satellite lasers

pioneered by Starlink for mesh networking at scale, creating coherent computing clusters across free space

instead of wired connections used in terrestrial data centers. Our existing Starlink constellation, with over

23,000 inter-satellite lasers, will be a crucial enabler of orbital AI compute, as its global network allows data

from our AI compute satellites in Sun-synchronous orbit to reach ground stations anywhere on Earth. The

SpaceX AI compute satellites will be designed for high rate, automated production to enable the scale of

satellites needed for the large amounts of compute planned in space.

There are material differences between connectivity satellites and AI compute satellites. Connectivity satellites

are primarily designed for communications, with substantial onboard equipment dedicated to phased-array

antennas, radio systems, and data transmission. In contrast, AI compute satellites are optimized for high-

performance computing. Key differences include significantly larger solar arrays to support higher power

requirements, substantially larger radiators for thermal management, different electronics centered on AI

accelerators rather than communications processors, and the removal of much of the communications hardware.

Our V3 satellite platform already incorporates proprietary chips, providing a strong foundation for the ability to

operate AI-focused electronics in space, and we expect to begin deploying our orbital AI compute satellites as

early as 2028.

The primary remaining challenge is one of scale. For example, a deployment rate of approximately 10 gigawatts

per year would require a materially lower manufacturing and launch cadence, which we believe would still

enable a commercially attractive AI compute business with strong economic returns. While our long-term vision

includes the ambition of deploying up to 100 gigawatts of power to orbit annually, which would require the

deployment of thousands of launches per year, assuming 100 kilowatts of compute power per metric ton and

Starship capacity to orbit of 100 metric tons, we believe we can be economically successful at significantly

more modest volumes.

153

Our 100 gigawatt annual power deployment goal is based on reasoned engineering analyses and design

parameters developed through our ongoing design and development work on next-generation AI compute

satellites. These analyses are based on currently available space-grade solar technology and do not require

fundamental technological advances beyond existing capabilities. Specifically, we expect these satellites to

leverage our already-designed V3 satellite platform. The core V3 satellite design is complete, and the AI

compute satellites are expected to generate substantially more power than V3 satellites. This performance is

expected to be achieved primarily through the use of significantly larger solar arrays. These satellites are

targeted to generate approximately 100 kW of compute power per ton, which initially will require

approximately five times the solar array output compared to V3 satellite designs.

We currently do not anticipate material supply constraints for space-based solar panels, as global production

capacity, including through our vertical integration efforts, is believed to be sufficient to meet its requirements.

We are actively developing the manufacturing, launch cadence, and operational capabilities that we believe

would be needed to support such launch rates.

The precise solar collection area, total system mass per satellite, and on-orbit assembly requirements associated

with this goal continue to be refined as part of our ongoing engineering efforts. In general, the approach

contemplates larger deployable solar arrays on each satellite, with no significant on-orbit assembly currently

anticipated.

•We will use our proven Starlink in-orbit technology to optimize our orbital AI compute. In order to

operate orbital AI compute satellites, we plan to build on our vast experience of operating approximately 9,600

Starlink broadband and mobile satellites in Low-Earth Orbit. In 2025 alone, Starlink satellites proactively

performed over 1,000 automated collision avoidance maneuvers per day guided by this technology to safely and

efficiently operate the constellation. This operating model gives us control over workload placement across

Earth and space while maintaining resilience through redundancy and fail safe systems. To ensure optimized

thermal management and power generation, we will design each satellite’s solar arrays to face the sun for

constant power while its housing radiator panels face cold deep space for radiative cooling. A high degree of

controllability will allow the satellite to be optimized for brightness mitigation, disposal, and other modes of

operation. As more advanced AI hardware becomes available, we plan to manage the lifecycle of deployed

systems by shifting older hardware to lower intensity workloads as performance characteristics evolve, and

retiring systems that are no longer needed through controlled end of life disposition, including transition to

graveyard orbits where appropriate. These retirements may occur sooner than our estimates for the useful lives

of our satellites, which estimates are based on engineering studies, historical on-orbit performance, propellant

life, utilization patterns, design enhancements across generations, and planned transitions to newer satellite

technology. Space based compute also introduces orbital debris risk, which we already manage at constellation

scale through our autonomous collision avoidance system across Starlink. To date, we have not experienced any

failures of our autonomous collision avoidance system that have resulted in satellite loss.

•We can manufacture our AI compute constellations at scale with rapid upgrade cycles. We have built one

of the largest satellite manufacturing operations in the world with standardized bus architectures, rapid iteration

cycles, and automotive-style production lines, enabling us to evolve bus architecture and subsystem design with

limited reliance on third-party suppliers. Our highly vertically integrated approach will be key to our mass-

scaling efforts and should allow us to deploy the latest AI processors. Our ability to quickly develop and deploy

new generations of AI compute to orbit will be a key advantage in maintaining frontier performance of the

constellation. We believe SpaceX will be the first and only company to manufacture satellites at the scale of

automotive manufacturing.

•We are building chip manufacturing capabilities to scale our access to AI compute hardware. We

announced a collaboration with Tesla in March 2026 to build the Terafab initiative with a long-term goal of

producing one terawatt of compute hardware each year. Intel joined the project in April 2026 and is expected to

contribute its expertise in designing, fabricating, and packaging ultra-high-performance chips to help Terafab

scale. In connection with such collaboration, we have agreed with Tesla on a general framework for the future

development of Terafab. Any specific projects undertaken pursuant to this framework will be subject to separate

negotiations and agreements (including any development timelines, milestones and capital expenditures) and

154

have not yet been determined. Our strategy for Terafab is to vertically integrate across design of lithography

masks, fabrication of logic and memory chips, design of advanced packaging and rapidly test and iterate in

order to improve chip design and performance.With this internal manufacturing capability, we plan to alleviate

potential future chip shortages at SpaceX, especially as we develop orbital AI at scale, and design chips that are

optimized for the space environment. We expect that our speed and cost advantage from vertical integration will

allow us to scale efficiently in AI chip manufacturing.

•We can leverage our terrestrial experience to build and operate compute clusters and AI workloads at

scale. We believe our experience operating compute infrastructure on Earth provides the technical and

operational foundation to extend these capabilities into orbit. For example, manufacturing and silicon defects in

AI processors can cause failures early in life. We plan to subject compute hardware to extensive pre-deployment

testing on Earth to identify early life failures before launch to reduce in-orbit disruption. Over time, we plan to

design AI compute processors optimized for the space environment. Our operating experience will be critical in

informing our orbital data center designs for highly reliable operations even with potential chip failures. This

capability is further supported by our flexible allocation of AI workloads across compute clusters, enabling us to

utilize orbital data centers for workloads without hardware reconfigurations or maintenance. For compute

hardware that does fail, we plan to leverage existing Starlink fleet management software to reallocate traffic to

other satellites and prevent cluster-level downtime. We further believe that our strong relationships with chip

makers enhance our ability to build a well-functioning, integrated AI compute system in space.

We Believe Our Infrastructure is a Distinct Advantage in Delivering Superior AI.We believe that the key

constraints in the continued growth of AI are physical – chip manufacturing, data center infrastructure, and power

generation; the future of AI will be determined by the control of the physical stack. We believe no other AI company

has better control over the full physical stack than SpaceX. We expect the combination of competitive cost per

token, our ability to deploy and operate data centers in orbit, and our strength in connectivity to result in more

scalable intelligence that is accessible globally at high speeds by way of the following structural advantages:

•Time to power. If we are able to deploy our AI compute satellite constellation, we believe it will enable

compute capacity to be deployed and expanded efficiently as capacity requirements grow. This approach will

also allow us to deploy new generations of compute hardware in quicker succession relative to terrestrial

approaches where data centers cannot be easily retrofitted for new compute hardware. Due to terrestrial

retrofitting limitations, adding terrestrial capacity typically demands building large, new data centers designed

for specific generations of compute hardware. This approach is usually burdened with long lead times for

activities such as power procurement, utility grid interconnections, and permitting before new computing

hardware can generate useful tokens. We believe our orbital, modular approach will allow us to circumvent

terrestrial power infrastructure constraints.

•Highly scalable compute capacity. Unlike terrestrial facilities constrained by physical footprint and

availability of power in a given location, orbital data centers leverage a decentralized mesh architecture. This

permits the aggregation of massive compute clusters interconnected over long distances by inter-satellite lasers

pioneered by Starlink. Space offers effectively unlimited power and vast expanse to sustain uninterrupted

operations as capacity grows. We believe this abundance of power and physical area will allow us to scale our

connected compute capacity faster and far beyond levels that are terrestrially viable.

•Low latency. Our satellite constellation provides a direct, orbital data path that circumvents the bottlenecks of

terrestrial communications networks. This architecture is particularly suitable to support high-speed

connectivity for latency-sensitive workloads, which we believe are increasingly valued in certain consumer- and

enterprise-facing applications.

•Global distribution. Because of the global coverage of our satellite constellation, not only can we deliver high-

speed, ultra-low latency AI solutions, we can do so anywhere in the world. We believe our increasingly global

network of Starlink satellites will enable us to deliver frontier intelligence, at high speed and reliability, to

communities and economies around the world.

155

Design and manufacture our own chips. Terafab aims to be the world’s largest chip manufacturing facility, with

the goal of achieving one terawatt of annual compute production capacity. While Terafab is intended to expand our

internal chip manufacturing capabilities, we expect to continue sourcing a significant portion of our compute

hardware from third-party suppliers. We view Terafab as complementary to these relationships, enabling us to

augment our access to compute hardware at massive scale and further complete our highly vertically integrated

compute platform by extending our control to the foundational chip layer. By developing end-to-end capabilities

spanning the design of lithography masks, fabrication of logic and memory chips, and advanced packaging, all in a

vertically integrated closed-loop single plant, we will be able to more rapidly iterate to improve chip design and

performance. We plan to design chips that are optimized for the space environment. This collaboration directly

enables our planned orders-of-magnitude increases in AI compute deployment in orbit which would be constrained

by pure reliance on external foundries. Leveraging shared engineering resources, intellectual property, and

infrastructure across Tesla and SpaceX, as well as Intel’s proposed contribution of its expertise in designing,

fabricating and packaging ultra-high-performance chips at scale, Terafab creates powerful ecosystem synergies that

accelerate innovation cycles and reduce costs. Just as we manufacture approximately 80% of Starship in-house,

enabling it to be the world’s most powerful and, eventually, the most cost-effective launch vehicle through full and

rapid reusability, we expect significant speed and cost advantages from Terafab’s vertical integration. We believe

this will provide us with a critical competitive advantage in the race to scale AI infrastructure, especially as we begin

our orbital AI compute satellite deployments.

Industry Overview

We are focused on three rapidly evolving industries: space, connectivity, and AI. Technological advancements and

breakthrough innovation are enabling what we believe is the next great economic frontier, as progress across space

launch, global communications, frontier models, AI compute, robotics, and automation reshape what is possible on

and off Earth. There are several key trends driving the growth and evolution of these industries in which we operate:

•Reusable launch and industrialized space operations are materially reducing the cost of access to orbit,

increasing mass carried per launch, and enabling high-cadence deployment of space-based infrastructure;

•High‑volume satellite manufacturing, combined with rapid constellation refresh cycles, is expanding the ability

for ubiquitous connectivity across unconnected, underconnected, and mobile “dead zone” areas; and

•AI, automation, and robotics are accelerating engineering iteration cycles, streamlining operations, and

revolutionizing complex construction, reducing reliance on scarce specialized labor while delivering faster,

more precise, and cost-optimized infrastructure.

The Space Industry

For most of the space age—dating back to the first launches in the 1950s—spaceflight was shaped by onerous

regulatory requirements and government budgets that determined launch cadence. The prevailing cost-plus

procurement model offered limited incentives to reduce costs or increase launch cadence, creating an operating

environment that constrained technological innovation. Government agencies served as the primary launch services

providers and the industry remained stagnant for decades. According to NASA, until the 2000s and the introduction

of the Falcon 9 rocket by SpaceX, global commercial launch activity averaged 25 to 35 launches per year. As a

result, the space industry remained a niche domain with limited ability to support large commercial markets or

scaled space-based infrastructure.

During this period, satellites—which comprised the majority of launch payload—were typically bespoke, expensive

systems requiring significant non-recurring engineering that consisted of development cycles that were measured in

decades. Launch vehicles were designed to be largely expendable and optimized for single-mission use, reinforcing

a low-throughput ecosystem that lacked flexibility, scalability, and responsiveness to evolving customer

requirements.

The need for more advanced launch capabilities became clear as space-based use cases expanded to include

communications, navigation, Earth observation, environmental monitoring, scientific research, Intelligence,

Surveillance and Reconnaissance, and access to the International Space Station. In 2006, NASA awarded SpaceX,

156

along with Rocketplane Kistler, the landmark Commercial Orbital Transportation Services contract that heralded the

age of commercial space launch, marking a shift toward a more scalable approach to accessing space. This inflection

point catalyzed a transition toward systems designed for more frequent operations, lower cost, and greater

operational flexibility.

Fundamental breakthroughs in high cadence, reliable, and affordable access to space—driven largely by SpaceX—

have expanded space from a purely mission-driven activity to a fully industrialized and commercial sector capable

of supporting and enabling industries far beyond traditional launch and satellites. SpaceX’s advancements reduced

the cost of access to orbit from tens of thousands of dollars per kilogram to just a few thousand dollars per kilogram.

Cost of Space Launches to Low-Earth Orbit

(constant 2021 $ per kilogram; plotted on a logarithmic axis)

As launch economics have changed rapidly over the last decade, demand for orbital infrastructure has expanded

dramatically. Commercial operators have launched thousands of satellites since 2015 as constellation architectures

scale and diversify. The number of active maneuverable satellites in orbit has grown from less than 1,000 in 2015 to

approximately 12,700 as of March 31, 2026. With approximately 9,600 Starlink broadband and mobile satellites in

Low-Earth Orbit as of March 31, 2026, SpaceX owns and operates approximately 75% of all active maneuverable

satellites. Additionally, launch activity has continued to grow, with approximately 220 metric tons of payload

launched to orbit in 2012 increasing to approximately 2,600 metric tons in 2025, of which over 80% was launched

by SpaceX.

Government demand is rising in parallel: according to the Space Foundation, excluding classified spending, U.S.

Government space spending in 2024 totaled approximately $77 billion. Notably, U.S. national security customers

have also awarded approximately $13.7 billion across the National Security Space Launch (“NSSL”) Program’s

Phase 3 Lane 2 contracts through 2032, supporting approximately 54 missions from 2025 to 2032, with the overall

Phase 3 manifest nearly doubling Phase 2’s manifest to 84 missions. Amid escalating geopolitical tensions that

further underscore the critical role of resilient launch infrastructure, we believe government space budgets around

the world are positioned for sustained, long‑term growth.

157

Falcon Heavy Boosters Landing

On the back of dramatically reduced launch cost pioneered by SpaceX over the past two decades, the global

economy is reorganizing around a new domain: space. We believe the development of a lunar economy will be

central to unlocking the full potential of this new domain and advancing the long-term transition to a multiplanetary

civilization.

The Connectivity Industry

Modern life relies on connectivity. Over the past several decades, the technologies that underpin global connectivity

have evolved rapidly, reshaping the way individuals, families, and organizations communicate, collaborate, and

access information.

Despite remarkable technological advancements, terrestrial networks remain constrained by the same inherent

structural limitations that have hindered them since their inception. According to the Global Satellite Operators

Association, terrestrial network infrastructure only covers approximately 20% of global land mass, resulting in

significant unserved and underserved regions across both developed and developing economies. This terrestrial

connectivity gap spans areas that are remote, difficult to build in, or economically impractical to serve—and also

includes mobile “dead zones” within otherwise well-connected areas and in urban markets. According to the J.D.

Power U.S. Wireless Network Quality Performance Study, U.S. wireless customers experienced service problems in

approximately one out of every 11 mobile interactions, even in well-connected areas. As demand for ubiquitous,

high-reliability connectivity continues to rise, terrestrial networks alone are increasingly unable to bridge the

widening gap between user demand and available coverage.

The development of large-scale LEO constellations represented a paradigm shift, breaking from the long-standing

dependence on terrestrial networks for global connectivity. Deployed at unprecedented scale—such as through

SpaceX’s Starlink and Mobile constellations—these satellites can provide high-speed, low-latency service that

integrates seamlessly with terrestrial infrastructure. This evolution has transformed satellite connectivity from a

solution of last resort into a core pillar of resilient, ubiquitous global communications.

158

Consumer Broadband

Residential internet access began with the dial-up connection in the late 1990s with maximum speeds of .056 Mbps,

when early users relied on narrowband copper phone lines to connect. As demand for speed and reliability grew,

dial-up gave way to DSL, cable, and eventually fiber, each increasing bandwidth and enabling more connected

devices. According to the Speedtest Global Index, the global average broadband download speed has increased to

approximately 120 Mbps. Satellite internet also emerged in the 1990s through geostationary orbit (GEO) systems

that extended coverage to remote and unconnected regions—beginning with early offerings such as Hughesnet’s

first satellite service DirecPC, which provided downstream speeds of roughly 400 kbps compared to dial-up

averages of 28.8 kbps—but these systems were constrained by limited throughput and high latency, making it

difficult to keep pace as consumer requirements evolved. Starlink satellites operate in Low-Earth Orbit, substantially

closer to the Earth’s surface than traditional geostationary communications satellites. This architecture reduces

signal latency and is designed to support broadband connectivity in remote and underserved areas. Each launch of

additional Starlink satellites increases the overall capacity of the network, which provides service globally.

In today’s digital landscape, consumers increasingly rely on seamless, high-performance connectivity to power all

aspects of connected life—from every day digital services to demanding applications that require high throughput,

consistent performance, and low latency. These needs are particularly challenging to satisfy in regions where

terrestrial networks are limited, degraded, or unavailable due to prohibitive deployment costs, rugged terrain, low

population density, or outdated infrastructure. Consequently, consumer broadband has evolved into a multifaceted

ecosystem, where diverse access technologies converge and providers compete based on superior reliability,

consistent performance, and an exceptional overall user experience.

Consumer demand for data is surging at a pace that terrestrial infrastructure has struggled to match. According to

International Data Corporation’s Global DataSphere, in 2025, global data generation was estimated to have reached

more than 585 exabytes of per day—up from approximately 10.8 exabytes per day in 2010—reflecting an immense

escalation in consumption. With fixed broadband connections projected to reach two billion by 2030 according to

Ericsson, and terrestrial expansion often economically unfeasible in remote and challenging regions, only space-

based systems can deliver truly global, ubiquitous, high-throughput coverage capable of supporting this explosive

growth in data demand.

Enterprise and Government Broadband

Enterprise broadband internet has evolved alongside residential internet, beginning with fixed private lines that

connected offices and infrastructure. As businesses adopted real-time, distributed workflows, they needed secure,

low-latency connectivity across multiple sites and mobile assets. Mobility became essential in sectors like

manufacturing, transportation, and logistics, extending connectivity demands beyond fixed locations into dynamic

environments that terrestrial networks often cannot support reliably or economically. Enterprises now expect

seamless, uninterrupted performance with instant failover where terrestrial systems are unavailable or unstable—

driving adoption of hybrid architectures that combine ground networks with space-based solutions.

Enterprise connectivity demand continues to rise as organizations digitize operations and rely on real‑time,

cloud‑based workflows that require secure, low‑latency connectivity across distributed sites and mobile

environments. This is particularly true in the case of aviation, maritime, and land mobility applications, where

aircraft, vessels, and ground fleets are inherently mobile and therefore unable to depend on continuous terrestrial

network coverage for connectivity. These platforms increasingly require resilient communications to support flight

and voyage operations, crew applications, passenger internet access, telematics, and port or shipboard logistics. In

aviation, legacy GEO-based systems that are still prevalent across most major commercial fleets typically provide

low Mbps speeds and significantly higher latency, often exceeding 500 milliseconds, falling well short of the

approximately 100 Mbps throughput and sub-50 milliseconds latency that today’s applications—such as streaming,

cloud services, and real-time collaboration—increasingly demand. Therefore, there is a need for modern LEO-

powered in-flight connectivity systems—such as Starlink Broadband—that can deliver passenger download speeds

exceeding 400 Mbps with latency as low as 21 milliseconds. Terrestrial networks cannot meet these evolving

demands where deployment is costly, complex, and slowed by regulatory constraints, and legacy satellite solutions

159

have not delivered the latency or consistency needed for enterprise‑grade applications, with average terrestrial ISP

download speeds at 120 Mbps and average latency from 7-34 milliseconds.

Defense and civil agencies similarly require secure, resilient, and global connectivity, often operating in contested or

infrastructure-poor regions where terrestrial networks are unavailable or vulnerable. As the battlefield becomes

increasingly connected, the need for robust, persistent connectivity across all domains is more urgent than ever.

Modern missions depend on high-throughput, low-latency connectivity for command and control, autonomous

systems, emergency response, and humanitarian operations, driving demand for architectures that maintain

performance where terrestrial systems fail. Substantial government investment into mission-critical, space-based

communication services illustrates the institutional reliance on LEO architecture for defense applications. High-

throughput, low-latency LEO constellations add a new architectural layer that enhances redundancy, operational

continuity, and flexibility across mission sets. There is an increasing need for purpose-built secure platforms—such

as Starshield, that can provide encrypted, high-assurance communications and modular payload integration—further

expand the utility of space-based connectivity for defense, civil, and national resilience needs. Together, these

advances position space as the foundational component of future mission‑critical communications architectures.

Satellite-to-Mobile Service

Since the early rise of mobile phones, terrestrial networks have expanded at immense cost and increasing density to

support successive generations of cellular technology—from the primarily voice-centric networks of the 1980s to

today’s high-speed 5G data networks. These investments have enabled much of the global population to become

well‑connected, yet the capital‑intensive nature of terrestrial build‑outs has resulted in vast geographic mobile “dead

zones” where coverage remains too expensive or is nonexistent. In many regions particularly those that are remote

or sparsely populated, extending towers is economically impractical for mobile network operators, resulting in large

segments of the population with limited or no access to reliable connectivity. Early satellite-based cellular options,

beginning in the 1980s with dedicated satellite phones, helped fill these gaps but required bulky hardware and

carried high usage cost, limiting them to narrow and mission-driven use cases. As consumer expectations for

ubiquitous coverage have grown, mobile network operators face structural limits in closing these “dead zones” with

terrestrial infrastructure alone, making LEO-based augmentation the most viable path to continuous, reliable mobile

connectivity at global scale.

Early satellite-to-mobile services (i.e., those connecting directly to standard smartphones) emerged in the 2020s with

support for basic messaging and, in some cases, voice in areas without terrestrial coverage. These offerings provided

more contiguous communication for safety, continuity, and remote operations. However, they were introduced at the

same time mobile data consumption was accelerating dramatically, and consumer expectations for “always-

connected” devices were rising. As a result, satellite-to-mobile technology is now evolving beyond emergency-only

communication. It is shifting toward enabling everyday smartphones to remain seamlessly connected when outside

traditional cellular or Wi-Fi range, integrating satellite connectivity into routine mobile usage, rather than treating it

as a contingency layer. At the same time, telecom operators have been reducing capital expenditures amid slower

revenue growth, weaker monetization, and declining returns on invested capital—pressures that have limited their

willingness to maintain historically high levels of network deployment. These shifts are also increasing demand for

harmonized, scalable spectrum allocations capable of supporting higher-capacity satellite-to-mobile services without

interfering with terrestrial networks, with the potential to add an incremental $1.4 trillion of economic growth over

the next 10 years, as forecasted by Cellular Telecommunications and Internet Association.

These industry shifts have opened the door for deeper collaboration among satellite operators, MNOs, carriers,

spectrum owners, device manufacturers, and regulators. As satellite network performance continues to improve and

these partnerships expand, satellite-to-mobile offerings—such as Starlink Mobile—are poised to evolve from a

“backup” layer into a meaningful complement to terrestrial networks, extending coverage and enhancing overall

network resilience and performance.

The AI Industry

Humanity is defined by our relentless pursuit of knowledge, with each transformative breakthrough dramatically

expanding our capacity to create, preserve, and share ideas across time and space. AI marks the next—and arguably

160

most consequential—chapter in this progression. For the first time, we are creating systems that do more than simply

amplify or transmit human-generated knowledge. These systems can reason, learn, and generate new knowledge

autonomously—synthesizing information, forming hypotheses, and in some domains even making original

discoveries. In doing so, they augment, accelerate, and will likely surpass unaided human cognition. This represents

a profound shift: we are moving from tools that simply extend the mind to autonomous agents and companions that

actively participate in the act of knowing.

Over the past decade, the convergence of big data, advances in AI hardware, and the breakthrough development of

LLMs have transformed AI from a speculative academic field into a foundational driver of the modern economy.

AI Compute

Massive demand for frontier AI models is accelerating the build-out of AI infrastructure at a pace and scale with few

historical precedents. Meeting projected AI needs will require $7 trillion in global data center investment through

2030, with generative AI workloads expected to account for roughly 70% of total data-center power demand by the

end of the decade. Each new generation of frontier models requires exponentially greater compute, following well-

established scaling laws that link model performance to the volume and quality of training data, parameter count,

and total compute expected. The rise of agentic AI and the potential emergence of artificial general intelligence are

expected to further amplify inference workloads, driving a step-function increase in compute requirements and the

corresponding data center capacity needed to support them. Frontier AI has become fundamentally infrastructure-

constrained. Only operators with access to massive amounts of power, very large GPU clusters and tightly integrated

training infrastructure can train cutting-edge models, and these systems exhibit non-linear performance advantages

that compound over time. Compute infrastructure scale helps determine model iteration speed, model quality, and

capital efficiency—making infrastructure itself a critical capability.

AI Frontier Models

A new class of frontier models has emerged, which includes LLMs and multimodal models. LLMs are neural

network-based models trained on massive datasets to interpret user questions and generate responses to highly

complex questions. LLMs can synthesize existing research, propose new ideas, and communicate in a natural

language that requires no programming expertise by the user. Demand for these tools has been explosive—according

to a YouGov survey, approximately 60% of Americans have used AI tools since December 2024, and 34% use AI

tools at least weekly. Multimodal models are AI systems that can process, understand, and generate outputs across

multiple types of data simultaneously—such as text, images, audio, video, and sometimes other modalities—rather

than being limited to just one (like text-only language models). Multimodal models offer several key benefits over

traditional unimodal (e.g., text-only) systems by processing and integrating multiple data types like text, images,

audio, video, and sometimes sensor data simultaneously. They provide richer contextual understanding, capturing

relationships and nuances across modalities that are invisible in isolation, leading to more accurate predictions and

reasoning.

AI frontier models are shaped by the values, objectives, and design choices of their creators. Model intelligence and

performance reflect decisions around data curation, training methodologies, alignment frameworks, and system

constraints, resulting in different reasoning styles, interpretations, and responses across models. Therefore, values

can be embedded in the technology, influencing accuracy, logic, and utility of the model outputs and how well

models can serve end users.

Following rapid frontier model innovation and broad adoption of chat-based tools, organizations are now beginning

to deploy agentic systems—AI that can use tools and operate with limited supervision. This marks the beginning of

what we believe will be a broader transition from co-pilots to agentic systems that enable high-complexity

workflows and create materially higher inference demand.

Consumer and Enterprise Applications

Advances in digital communication have reshaped how information is created, shared, and consumed, laying the

foundation for today’s social media platforms. These platforms have become essential channels for digital

advertising by combining large‑scale user engagement with targeted content and ad distribution. Recent advances in

161

AI are further strengthening advertising, allowing enterprises to optimize campaigns and measure outcomes. At the

same time, consumer expectations for AI‑powered tools are rising, with users seeking timely, accurate and

trustworthy information across an expanding universe of digital content.

We believe the ongoing convergence of consumer platforms, consumer AI, and integrated digital services will

accelerate the emergence of super‑app ecosystems that combine communication, content creation, information,

commerce, and banking within a single platform. These trends are expected to expand the role of internet platforms

as distribution channels and support next‑generation AI‑enabled applications and advertising solutions.

For enterprises and governments, frontier models and agentic AI—autonomous systems capable of multi-step

reasoning and independent task execution—are beginning to manage increasingly complex processes and

workflows. As of February 2026, more than 80% of Fortune 500 companies were using AI active agents. Entire

industries are being reshaped by AI-driven applications, including agentic commerce (personalized AI-directed

shopping), vibe coding (software development with minimal or no human-written code), and autonomous driving for

vehicles.

The ultimate frontier in AI is human augmentation: creating systems that amplify and multiply human reasoning,

creativity, decision-making, and productivity, enabling people to perform highly complex tasks with unprecedented

speed, scale, and insight. By enhancing how humans think, learn, and interact, such systems act as cognitive

multipliers, supercharging individual and collective capabilities far beyond biological limits. As AI evolves, we

expect both consumer platforms and enterprises to adopt increasingly agentic systems that serve as powerful

extensions of human intelligence. These tools will orchestrate multi-step workflows, interact seamlessly with

business applications, and accelerate operational processes, with humans at the center of judgment, creativity, and

strategy. Emerging efforts in enterprise AI illustrate how future systems could coordinate entire business functions

as force multipliers—dramatically expanding what a human team can achieve with minimal scaling friction and

maximal leverage. Human augmentation also offers a transformative solution to the escalating effort required for

breakthroughs in technology and beyond. For example, the human effort needed to sustain Moore’s Law (chip

density doubling approximately every two years) has increased eighteenfold since the early 1970s; AI augmentation

could reverse this trend by empowering engineers, researchers, and innovators to iterate faster, explore more

possibilities, and achieve exponential progress with smaller, core teams of experts.

As humanity expands beyond Earth, augmented human intelligence will be essential to managing the immense

operational, scientific, and logistical complexity of a spacefaring civilization. The core promise of augmentation lies

in multiplication: AI not as a substitute for human minds, but as an amplifier for human ingenuity, curiosity and

purpose that unlocks new frontiers of what humans can accomplish together.

Our Strengths

We have an intense, mission-driven, and engineering-first culture that seeks to achieve what many have deemed

impossible. We make the incredible and extraordinary possible and repeatable by continuously leveraging our core

strengths:

Global Leadership in Orbital Launch Services

Our unique ability to reliably, quickly, and cost efficiently launch rockets at scale into space is our core competitive

advantage that enables other parts of our business. Our launch capabilities form the foundation of our orbital

infrastructure and have created new multi-trillion-dollar opportunities in space, global connectivity, and AI. We

believe no other launch provider is competitive at this scale today, nor is likely to become so in the near term. Our

fleet of 24 flight-proven, reusable rockets and our growing share of total mass delivered to orbit has increased every

year since 2021. Reusability completely changes the economics of space access. Qualified for 40 launches, our

reusable rockets can fly multiple times with only minimal refurbishment between missions, sharply lowering the

cost per launch, while boosting our launch rate, asset use, and overall efficiency compared to traditional expendable

rockets. As a result, we can offer competitive launch prices, rapidly deploy our own satellites and infrastructure, and

make it easier and cheaper for us to pursue new opportunities requiring orbital access. Our higher launch rates and

reusability also create a virtuous cycle: more flights lead to faster improvements in design, manufacturing, and

operations through accumulated experience. Additionally, not only did we demonstrate at least a 10-year advantage

162

over the rest of the industry when we first landed our Falcon 9 booster back from space in 2015, but we have

continued to invest significantly in further increasing our lead by pursuing full and rapid reusability at scale,

including investing over $15 billion in our next-generation rocket, Starship.

Unrivaled Satellite and Connectivity Platform across Design, Manufacturing, Deployment, and Operations

We are able to design, engineer, and manufacture the world’s most advanced satellites at scale, enabling the creation

and scaling of new businesses leveraging this core satellite technology platform, including: Starlink Broadband, our

space-based internet broadband service; Starlink Mobile, our global satellite-to-mobile service; and emerging AI

initiatives. Unlike traditional satellite manufacturers that rely on fragmented supply chains and low-volume

production, we have built an integrated satellite platform that spans architecture, chip design, software, power

systems, and final assembly. As we rapidly iterate on our next-generation satellites in-house, some others are

contracting outsourced manufacturers to build satellite architectures with capacity comparable to satellites that we

retired years ago. As of March 31, 2026, our constellation also incorporates over 23,000 inter-satellite lasers that

create a dynamic mesh network in space, enabling data traffic to route through orbit rather than relying solely on

terrestrial backhaul infrastructure. By controlling satellite design, production, launch and operations, we can tailor

payloads, networking capabilities, and power requirements to support new use cases. For example, our AI compute

constellations will leverage our core satellite technologies already developed for our existing Starlink constellations.

We will build new satellites that can host processors for high-density compute payloads, offer enhanced power

generation with larger solar panels and storage systems, and enable higher-capacity networking capabilities to

support low-latency workloads in orbit. Our high-throughput manufacturing capabilities—combined with our launch

capabilities—enable us to produce and deploy thousands of satellites per year, an uneconomic proposition for those

lacking an ability to deliver substantial mass into space. This capability accelerates our deployment timelines and

allows us to commercialize entire constellations with capital efficiency that we believe is difficult to replicate.

Our global connectivity platform, Starlink, is powered by the world’s largest LEO constellation and supported by

our vertically integrated launch and satellite manufacturing capabilities to enable the delivery of high-speed, low-

latency broadband and mobile connectivity to homes and businesses everywhere in the world. Our vertically

integrated model allows us to provide reliable service with unmatched speed and cost across geographies where

traditional terrestrial infrastructure has been limited, uneconomical, or unavailable.

Truth-Seeking AI Model Enhanced by Real-Time Data

AI frontier models are shaped by the values, objectives, and design choices of their creators that influence accuracy,

logic, and utility of the model outputs. We believe Grok represents a differentiated approach to AI, grounded in a

core objective of truth seeking and powered by continuous, proprietary access to real-time data inflows through its

integration with X. With approximately 350 million daily posts, X enables freshness, relevance, and contextual

awareness for Grok that we believe is a competitive differentiator. This direct, real-time access to the information

and human discourse on X enhances Grok’s truth-seeking capabilities by grounding outputs in up-to-date knowledge

and diverse viewpoints.

This architecture reflects our core philosophy that maximizing truth seeking—through the active, relentless pursuit

of what is objectively true about reality, grounded in evidence, logic, empirical data, and first principles thinking—

drives superior model outputs and higher utility intelligence. By combining our unique truth-seeking model with

proprietary access to one of the world’s largest real-time information platforms, we believe Grok can deliver the

most objective and relevant insights and best serve high-frequency, high-value use cases across consumer and

enterprise AI applications.

Extreme Vertical Integration Enabling High Velocity and Superior Cost Efficiency at Scale

While conventional aerospace manufacturing relies heavily on fragmented and outsourced supply chains, we operate

with extreme vertical integration. By designing and manufacturing a significant portion of our components in-house,

we bypass many of the slow, bloated sourcing channels that structurally constrain the rest of the industry. For

example, approximately 80% of Starship, SpaceX’s next-generation launch vehicle, is manufactured in-house. Our

vertical integration allows us to achieve iterative cycles in weeks, compared to years for some legacy companies,

enabling us to build newer, more technologically advanced products faster than many of our competitors. We

163

believe this technological and logistical gap is widening meaningfully as our speed and cost advantage compound.

Our vertical integration extends beyond design and manufacturing—it permeates our entire business model,

encompassing engineering, deployment, and operations. We are the only company building integrated hardware and

software infrastructure of the future across space, connectivity and AI. This end-to-end control allows us to deliver

value through structural advantages in speed, cost and quality.

Our strong belief in the benefits of extreme vertical integration is further exemplified by our acquisition of xAI. We

not only develop best-in-class models to support the application layer of AI, where we leverage real-time data

ingestion from X (subject to some limitations for certain content), but we also own and operate the physical compute

infrastructure required to train and run inference on those models, providing a substantial cost and speed advantage.

Through our Terafab initiative together with Tesla and Intel, we intend to further extend our vertical integration to

chip design and manufacturing to alleviate potential future chip shortages at SpaceX, optimize compute

performance, and reduce overall compute costs. Intel will contribute its expertise in designing, fabricating, and

packaging ultra-high-performance chips to help Terafab scale. This highly vertically integrated approach allows us

to train and iterate our frontier models at high velocity, accelerating development cycles, eliminating external

bottlenecks, and driving rapid, continuous improvements in model performance. Compute availability is also critical

for running more complex workloads and delivering higher performance inference at scale. As AI adoption

accelerates and demand for low-latency, high-throughput inference increases, we believe operators with the ability

to support and efficiently allocate compute across both training and inference workloads are best positioned to win

the AI race. Our human augmentation solutions are being designed to capitalize on this shift, enabling us to deliver

superior performance for our customers. This advantage of vertical integration exists in both a terrestrial context,

where we own our own data centers and the associated power infrastructure, and eventually in a space-based

context, where we are planning to build our own orbital AI compute infrastructure. The key constraints in the

continued growth of AI are physical—chip manufacturing, data center infrastructure, and power generation.

Differentiation is rapidly shifting from model architecture alone to AI compute scale, cost efficiency, power

availability, and speed of deployment. We believe that physical infrastructure, not models, will be the primary

competitive differentiator for AI companies, and no other AI company has better control over the full physical

infrastructure than SpaceX.

Unique Ability to Scale New Trillion-Dollar Markets Across Space, Connectivity, and AI

We believe space represents the largest economic frontier in human history. We believe we have a distinct ability to

identify, activate, and commercialize new multi-trillion-dollar markets that did not previously exist. Historically,

space access was impaired by high launch costs, low flight cadence, and limited demand. While such constraints

may limit others’ ability to access space at a scale, our ability to build large-scale and complex hardware

infrastructure is a meaningful competitive advantage. By pioneering the world’s first and only fleet of reusable

rockets at scale, we revolutionized space access through dramatically lower cost and unmatched reliability.

Lowering costs by orders of magnitude does not just expand the launch market, it enables the creation of entirely

new industries on Earth and in space that have historically been technologically and economically infeasible for

others to access historically.

When we have identified a new trillion-dollar market opportunity to pursue, we design a solution rooted in the same

world-class engineering and first-principles thinking that has driven our technological breakthroughs and success to

date. Our first trillion-dollar market was connectivity: we founded Starlink, a satellite service supported by our low-

latency, high-speed LEO constellation. Starlink required the rapid, low-cost deployment of millions of kilograms of

hardware into orbit, a feat economically impossible to solve for anyone lacking our foundational launch capabilities.

Our Starlink constellation powers a global connectivity platform capable of supporting the world’s largest and most

advanced space-based internet broadband service and satellite-to-mobile service, enabling high-speed internet access

to homes, enterprises, governments, and mobile users around the world. We believe our next trillion-dollar market is

AI compute, and we expect to leverage our rockets and satellites for massive orbital deployments of AI

infrastructure. We believe this AI compute infrastructure will help us develop and monetize the Grok model faster

than other AI companies that are dependent on finite sources of power on Earth. No other company has built the

capabilities to create value across all these end markets at scale.

164

In addition, we believe we are poised to catalyze transformative breakthroughs in other industries on Earth and in

space such as long haul point-to-point terrestrial travel, in-orbit manufacturing, passenger and cargo transportation to

the Moon and Mars, manufacturing and energy production on the Moon and Mars, and asteroid mining. In

particular, we believe that if we achieve our goal of establishing a lunar presence, it will potentially enable terawatt-

scale annual AI compute growth, support deeper space exploration and industrialization, and serve as a stepping

stone to establishing a civilization on Mars. As we continue to scale and expand into new trillion-dollar markets, we

expect our more mature businesses will continue to generate substantial cash flows, enabling us to reinvest in

emerging opportunities.

Business Models that Are Incredibly Difficult to Replicate

Our business model is simple to describe: leverage our unparalleled launch capabilities to reduce the cost of access

to space, apply first-principles thinking and world-class engineering to solve large structural constraints, vertically

integrate across the value chain, continuously improve cost efficiency and throughput, and reinvest cash flow to

expand our capabilities and create new markets. While simple to describe, we believe this model is extraordinarily

difficult to replicate. We believe no other organization can execute this combination of reusable orbital launch

systems at industrial scale, breakthrough engineering designs with reliable high-volume manufacturing, full stack

proprietary software, and end-to-end operational control. These capabilities reinforce each other, and our vertical

integration enables faster innovation cycles and structural cost advantages that widen our competitive advantage.

Our business model has allowed us to build a diversified portfolio of complementary businesses and revenue streams

from a common technological foundation. Our Space segment generates revenue from commercial and government

customers, while also serving as the backbone for our Connectivity segment which generates highly predictable and

recurring subscription revenue from Starlink broadband consumer, enterprise, and government customers, as well as

Starlink Mobile subscribers. The result is a powerful, self-reinforcing value creation cycle: success in one business

fuels faster growth in the others, enabling reinvestment into the next frontier. We believe this model has the potential

to create compounding value across our ecosystem, allowing our lead to grow and become more durable over time.

Mission-Driven Culture and World-Class Talent

We have the benefit of being founded and led by Elon Musk, one of the great visionaries of our generation. We

believe that our ability to attract and retain world-class technical and engineering talent is a significant competitive

advantage. Our founding goal of making life multiplanetary serves as the ultimate mission-driven filter and retention

tool, which has only been enhanced by xAI’s truth-seeking mission of understanding the universe. Top engineers are

drawn to SpaceX to work on some of the hardest, most consequential problems facing humanity—doing things that

have never been done before, like landing and re-using rockets, working towards making humanity multiplanetary,

and gaining a better understanding of the mysteries of the universe through AI. They are also drawn to our intense,

engineering-led, first-principles culture, which treats the laws of physics as the only true constraints. We reinforce

this culture through “The Algorithm,” a five-step iterative process that emphasizes making the requirements less

dumb, deleting unnecessary processes or parts (embracing the principle that the best part is no part), only then

optimizing what remains, accelerating cycle time, and automating only proven processes. Our organizational

philosophy embraces failure as an essential learning opportunity and maintains a relentless focus on efficiency and

speed, enabling rapid iteration and repeatable execution on the hardest technical problems. To this end, our

engineering-oriented organization maintains access to some of the world’s most selective talent pool. In 2025, we

accepted under 2% of our engineering applicants, reflecting our ability to be highly selective and hire among the

best talent in the industry. We also foster commitment by aligning employee interests with organizational success:

our broad-based employee ownership program ensures that those who help us build the future are also direct

beneficiaries of our success. This commitment to quality and mission results in exceptional employee loyalty,

reflected by an average tenure across our broader SpaceX leadership team of 12 years.

Our Growth Strategies

We have created what we believe to be the world’s most ambitious vertically integrated innovation engine that

captures significant growth across three domains: Space, Connectivity, and AI. While our Space segment provides

us with a foundational competitive advantage that enables all other parts of our business, our Connectivity and AI

165

segments are expected to be the primary driver of revenue growth in the near term. In the next few years, we are

focused on increasing the monetization of our existing Connectivity infrastructure and our existing AI user base. We

also intend to continue to build out our AI infrastructure, which we expect to enable growth as we address the

significant AI market opportunity. Our growth strategy aligns with our value creation cycle where we identify

emerging opportunities, invest in innovation, rigorously test and iterate, launch new offerings, and generate strong

cash flows to fuel the next wave of breakthroughs.

Space

Increase launch payload capacity. We plan to drive meaningful growth in payload delivered to orbit (mass to orbit)

through higher launch cadence and increased payload per launch, while enhancing launch efficiency and reducing

costs. Our next-generation fully and rapidly reusable Starship V3 vehicle is designed to carry 100 metric tons to

Earth’s orbit in a reusable configuration, driving substantial improvements in payload capacity per launch, while

enabling significantly more frequent flights, at unparalleled cost efficiency. To date, we have executed 11 Starship

flight tests. We have also scheduled a 12th flight test, which will debut the next generation Starship vehicle and

Super Heavy booster, powered by the next evolution of our Raptor engine and launching from a newly designed pad

at Starbase. We expect Starship to commence payload delivery to orbit in the second half of 2026. We have

achieved innovative milestones, such as the creation of booster catches using “chopstick” arms that facilitate rapid

refurbishment and reuse, including launching multiple times per day. To enable a more frequent launch cadence and

overall greater payload delivery, we are also expanding our ground launch infrastructure, including investing in

additional pads, on-site propellant production, and other support facilities, and investing in future generations of

Starship, which could carry 200 metric tons in capacity, potentially as soon as Starship V4. We expect these efforts

to continue to drive launch payload growth that is expected to provide the foundational capacity needed to scale our

Starlink Broadband and Starlink Mobile constellations that underpin our Connectivity platform. Our growing

payload capacity is also intended to underpin the deployment of orbital AI compute that will accelerate our AI

business, as well as benefit third-party customers who use our launch offerings.

Establish the lunar economy. Advancing access to the lunar surface represents an important next step in the

evolution of our Space segment and is a prerequisite for long-term commercialization beyond Earth. We are focused

on developing the capability to transport significant amounts of cargo and crew to the lunar surface in a repeatable

and economically viable manner. We believe this capability will also enable creating a petawatt-scale AI

constellation through the use of lunar satellite production and a lunar mass driver for launch activities. By leveraging

Starship’s expected fully and rapidly reusable capabilities and in‑space refueling, we expect to materially reduce the

cost of lunar missions relative to historical norms. Our initial efforts will prioritize lunar cargo landings and

returning Americans to the Moon, followed by expanded crewed missions that we believe can establish a continuous

flow of cargo and humans between Earth and the lunar surface.

We believe that the foundation of a commercial lunar economy begins with achieving infrastructure development,

lunar resource utilization, and high bandwidth communications at scale. This requires the ability to mine, extract and

process raw material for the production of solar power on the lunar surface. Combined with the ability to locally

produce water and fuel, we believe these capabilities would enable sustained lunar operations, support lunar

exploration, and provide the foundation for humanity’s permanent presence on the Moon. The lunar base would then

allow sustained, high volume testing of new technologies in a space environment much closer to Earth than deep

space.

We intend to establish lunar‑based manufacturing capabilities, including factories to produce large‑scale AI compute

satellites. We believe we can efficiently launch our satellites at scale, namely due to the potential use of a lunar mass

driver that is capable of high-frequency, low-cost launches of satellites from the lunar surface. By shifting energy

and material and mass-intensive satellite and solar manufacturing activities off Earth that leverage sustainable power

generation and the Moon’s low gravity, we aim to significantly reduce costs and terrestrial resource constraints. We

expect to use raw materials from the Moon to construct most of the mass of the satellites and ship chips and other

lower mass elements from Earth. This roadmap positions the Moon not only as a potential gateway to Mars and

space exploration, but as the first space-based industrial economy at scale.

166

Once resource utilization capabilities are proven feasible, we believe there is an opportunity to commercialize the

harvesting and exportation of rare materials, which is estimated to be present on the Moon in quantities exceeding

one million tons and has potential applications in future nuclear energy and quantum computing systems. Large-

scale access to these resources, coupled with the Moon’s low gravity, could unlock the potential for scalable growth

by establishing a vertically-integrated resource extraction, processing and exportation hub. Using Starship’s high

payload capacity, we believe these materials could be economically transported directly to Earth. In parallel, the

Moon could function as a proving ground for closed-loop ecosystems, long-duration habitats, and autonomous

construction techniques, all of which are essential for industrialization. Over time, this infrastructure has the

potential to position the Moon as a strategic industrial and transportation node.

Establishing lunar operations for mining, refueling, manufacturing, and habitation is subject to a variety of

interconnected engineering and other hurdles as well as known and currently unknown risks and uncertainties. These

include hurdles, risks and uncertainties that relate to, among other things, transporting and deploying heavy

equipment to the lunar surface, developing reliable power generation and storage systems, extracting and processing

lunar resources at commercial scale, operating equipment in extreme temperature, radiation and dust conditions,

maintaining communications and navigation infrastructure, and supporting long-duration human presence in a

remote and hazardous environment.

Connectivity

Grow Starlink Broadband customers.In the near term, we are focused on increasing global awareness of our

Starlink brand and capabilities to grow our base of Starlink Broadband subscribers and to increase Starlink

Broadband adoption in new and existing markets.

•Starlink Consumer Broadband. We have grown the number of Starlink Subscribers rapidly over the last

several years. As of March 31, 2026, we had approximately 10.3 million Starlink Subscribers across 164

countries, territories, and other markets. These subscribers represent a small fraction of the estimated 3.3 billion

potential end users in the markets we currently serve, many of whom still lack reliable high-speed broadband.

Because we report Starlink Subscribers on a per‑Service Line basis, the number of individual end users who

access Starlink is already likelymeaningfully higher than 10.3 million, as multiple people may share a single

Service Line, including within a household. We intend to grow the number of Starlink Subscribers by

expanding our consumer distribution network across thousands of authorized retail stores globally and execute

region-specific marketing campaigns to increase Starlink brand awareness. By clearly demonstrating Starlink’s

superior speed, low-latency, affordability, and ease of installation—not only in rural, remote, and infrastructure-

limited areas, but also in suburban and urban areas with wireline broadband options—we expect to drive

meaningful subscriber and revenue growth.

•Enterprise and Government Starlink Customers. We plan to drive growth in enterprise and government

Starlink customers through our direct, vertical-specific sales model. In recent years, we have assembled

dedicated sales and engineering teams to market and support fleet-wide conversions in the aviation and

maritime sectors. This has enabled partnerships with many of the world’s leading airlines, including United

Airlines, Southwest Airlines, Qatar Airways, Lufthansa Group, British Airways, Alaska Airlines, and Hawaiian

Airlines, many of which have implemented or committed to fleet-wide Starlink installations for seamless in-

flight connectivity. We have also partnered with premier cruise operators, such as Carnival Corporation, Royal

Caribbean Group, MSC Cruises, and Norwegian Cruise Line Holdings, for full-fleet deployments that deliver

reliable high-speed internet across thousands of vessels worldwide. In addition, we have partnered with land

mobility operators, including John Deere and the California Fire Department, as well as passenger rail operators

such as Brightline (Florida), and Italo Treno, to provide remote monitoring and management of their fleets. We

are actively driving growth in these sectors by onboarding new major airlines, cruise lines, and land mobility

operators around the world, expanding existing relationships through deeper fleet penetration, and introducing

advanced service tiers to make Starlink the standard connectivity solution for aviation, maritime, and land

mobility customers globally. We also intend to expand our government customer base, securing major contracts

with the United States and allied governments while delivering secure, resilient, and mission-critical

connectivity for defense operations, humanitarian efforts, disaster response, and national security applications in

even the most remote and challenging environments. We also serve a broad fixed‑site customer base across

167

industries such as retail and financial services that require high availability for critical operations as well as

reliable connectivity in remote or hard-to-serve locations. As companies continue to invest in secure and

resilient networks to keep critical infrastructure—such as point‑of‑sale and payment processing systems—we

see an opportunity to grow our broad fixed‑site customer base, often starting as back-up and then transitioning

to primary.

Expand our Starlink Mobile offering.As of March 31, 2026, we provide Starlink Mobile services to approximately

7.4 million monthly unique devices across approximately 30 countries. We partner with leading device

manufacturers, application developers, and mobile network operators to enhance the services we provide over one

satellite network, including over-the-top voice, video, and messaging. In 2025, we entered into agreements to

acquire 65 MHz of spectrum in the United States and certain global Mobile Satellite Service spectrum licenses from

EchoStar, which will enable a step-change in the possibilities for our Starlink Mobile service. Furthermore, we

anticipate that Starship will be able to deploy approximately 50 mobile satellites per launch, significantly increasing

capacity per launch and accelerating the deployment of our next-generation constellation. With the deployment of

our next-generation constellation, which is designed to fully utilize the acquired spectrum, and the expansion of our

MNO partnerships, we aim to further deliver on our goal of providing connectivity for everyone and substantially

reducing mobile “dead zones” worldwide—eventually with 5G connectivity to unmodified cell phones and IoT

devices globally.

Increase the capacity of our constellations. Our current constellations of approximately 9,600 Starlink broadband

and mobile satellites, including over 3,000 satellites deployed in 2025, support over 700 Tbps of cumulative

downlink capacity. To support larger numbers of customers through our Connectivity segment, we plan to materially

increase the capacity of our broadband and mobile constellations. For our Starlink broadband constellation, we will

continue deployment of more of our V2 Mini satellites, and in the second half of 2026, we expect to begin

deployment of our next-generation V3 satellites, each of which is designed to offer one Tbps of downlink capacity

per satellite. We expect Starship will be able to deploy up to 60 V3 satellites per launch, representing a twenty-fold

increase in downlink capacity deployed per launch compared to Falcon 9, enabling a more rapid expansion of our

Starlink broadband constellation at a significantly lower cost. For our Starlink Mobile constellation, we currently

have approximately 650 existing dedicated mobile satellites. We are developing more comprehensive satellite-to

mobile services, which we refer to as our Starlink Mobile Gen2 services, including broadband data and IoT

connectivity, which are expected to deliver resilient, infrastructure-independent connectivity worldwide and enable

5G connectivity.

We plan to expand our mobile constellation by deploying our next-generation mobile V2 Mobile satellites in 2027

which, combined with the EchoStar spectrum acquisition and optimized 5G protocols, are expected to increase

capacity by orders of magnitude compared to our first-generation constellation. In the U.S., the FCC approved the

EchoStar license transfer in May 2026, and we separately expect to receive the remaining necessary U.S. regulatory

authorizations in the second or third quarter of 2026. While these authorizations would be sufficient from a U.S.

regulatory perspective, we still require our V2 Mobile satellites to be in orbit and must complete the acquisition of

the relevant spectrum from EchoStar before we can commence our planned commercial Gen2 service in the United

States. Internationally, we have filed applications in nearly every country in which we intend to operate our Gen2

service, and approvals have been granted in a limited number of these jurisdictions to date. Each international

jurisdiction presents its own regulatory process and timeline, and we cannot predict when or whether approvals will

be granted in any given market. In addition, our Gen2 service is subject to ITU coordination requirements. We have

an operational coordination agreement with EchoStar, which we expect to continue through 2026 and 2027, under

which EchoStar has agreed to protect our lower-priority S-band V2 Mobile constellation. By prioritizing these step-

change capacity increases in our satellite-to-mobile capabilities, we expect to both enhance high-speed, low-latency

service quality in existing markets and provide services to previously capacity-limited and unserved regions,

including dense urban areas and emerging markets.

AI

Grow consumer AI platform monetization. We plan to continue to grow revenue from our AI platform, the Grok

application, by increasing monetization of our existing user base. We will leverage our unique combination of

real‑time data, large‑scale distribution, leading foundational model, and hardware expertise to increase the number

168

of Grok subscribers. Subscribers benefit from enhanced functionality, exclusive features, and access to our latest AI

models. Since the introduction of our Grok subscription offering in 2025, we have increased the number of available

features to add value to our subscribers, including providing access to our latest and enhanced AI tools. We plan to

continue adding new features and functionality while releasing increasingly capable Grok models to increase the

penetration rate of our subscriber base. Our AI segment has demonstrated exceptional model velocity: since

launching Grok, we have developed leading frontier models at a far faster rate of innovation than others. We

continue to invest in scaling Grok through subsequent generations, including Grok 5. Our roadmap for future models

contains multi-trillion parameter models, which could represent a step change in reasoning depth and overall

intelligence. We believe this pace of innovation strengthens the value proposition of our subscription offerings and

supports long term subscriber growth. While our subscriber growth has been strong, we believe we are still early in

increasing paid penetration across our Grok user base. We further believe there might be an incremental

monetization opportunity by introducing advertising into our stand-alone Grok offering.

Grow X monetization. We intend to drive X revenue growth by increasing engagement across our users, increasing

X Premium subscriber conversion, growing advertising revenue per user, and diversifying our advertising base. We

continue to evolve X into an “Everything App,” integrating real-time information, communications, media,

payments, banking, and more within one consumer app experience. This can improve the usefulness of X, and

therefore increase the usage and monetization potential of X. We have demonstrated rapid product launch velocity,

with frequent features and products launched since 2023, including Grok integration, long‑form video, audio and

video calling, secure messaging, tool calling, long-form articles, and creator tools. We plan to further broaden the

value proposition of X through offerings like Money, a product we launched in beta in November 2025, which aims

to expand platform utility by enabling payments and other financial services. We updated X chat in 2025, featuring

end-to-end encryption and no connection to our ad personalization, unlike other messaging services. We intend to

further embed Grok throughout X to enhance discovery, analysis of posts, user support, and personalization,

increasing the usefulness of X and further improving the value of a paid subscription.

We also expect to grow advertising revenue per user and to diversify our advertiser base over time because of X’s

compelling advertiser value proposition—large-scale user engagement, real-time content, and advanced AI-driven

performance marketing tools. We intend to drive further advertising revenue growth by improving our performance

advertising capabilities, embedding AI to optimize ad campaigns, and launching richer ad formats, including those

that increase advertiser return on ad spending and their spend with us. In determining our advertising rates, we use

an auction process in which advertisers bid to have their ads shown to the audience they are targeting, except for

certain reserved inventory, which is sold on a fixed price basis. We provide advertisers with several engagement

metrics, including: the number of impressions, price per ad, clicks, and conversions. Currently, Grok API access is

not included in our advertising rates to advertisers. We do not currently sell or offer advertisers the ability to place

ads on the Grok API.

We also expect X’s real-time content stream and engagement feedback, subject to some limitations for certain

content, to strengthen our advertising product performance and relevance, improving outcomes for both consumers

and advertisers, and increasing retention. We also began a phased roll-out of our new advertising platform, including

the new X Ads Manager, in April 2026. X Ads Manager is designed to help advertisers launch better campaigns

faster, with AI-powered systems enabling more precise, relevant, and dynamic ad delivery and a centralized

workflow for campaign creation, optimization, and real-time monitoring. Grok supports this strategy by helping

advertisers with campaign creation, creative optimization, and alignment with trending topics and user intent.

Deepen enterprise and government adoption. We believe adoption of AI by both enterprise and government reflects

a structural industry shift, with room for substantial long-term growth. Our Grok Business, Grok Enterprise, and xAI

Gov offerings position us to scale in tandem with broader enterprise and governmental AI adoption. Our Grok API

further extends our reach by enabling developers to integrate our models directly into their applications and

workflows. We intend to further support our enterprise offerings with a specialized salesforce and forward deployed

engineers, engineers who embed directly with a client to implement our solution, to support customer acquisition

and expansion.

Increase the scale of our terrestrial power and AI compute infrastructure. We plan to rapidly scale our terrestrial

AI compute infrastructure through the continued deployment of large-scale clusters to support the training and

169

inference of our AI models. To rapidly bring gigawatt-scale data centers online, we leverage world-class

engineering, first-principles thinking and deep “shovels-to-tokens” vertical integration. Our AI compute facilities,

COLOSSUS and COLOSSUS II, collectively provide approximately 1.0 gigawatt of compute power, with additional

power capacity available for data center operations. COLOSSUS II will also provide the compute to train our next-

generation Grok 5 AI model. We expect that once fully operational, the next phase of expansion at COLOSSUS II

will represent an additional 400MW of compute capacity. Our first-principles thinking enables us to build coherent

compute at scale and at rapid speed with lower costs than most other companies in the industry. We brought the first

cluster of COLOSSUS online in 122 days, repurposing the shell of an existing factory, and the first cluster of

COLOSSUS II online even faster in 91 days. As an illustrative comparison, an industry benchmark to bring online a

100 megawatt greenfield data center is approximately two years. We also demonstrated a significant improvement in

cost efficiency, achieving data center construction costs for COLOSSUS II that are considerably lower than industry

benchmarks on a per megawatt basis. As AI workloads increase in complexity and scale, data center operators face

constraints related to power density, cooling, network bandwidth, supply chain management, construction expertise

and capital deployment. Our experience in designing mission-critical hardware systems, optimizing power

efficiency, and operating distributed infrastructure networks provides a differentiated foundation for continuing to

grow and advance the next-generation compute platform. We believe that continued investment in our compute

infrastructure is critical to supporting long-term consumer and enterprise growth as AI adoption accelerates, while

also providing a powerful foundation for our transition to orbital AI compute at scale.

In addition, our leadership in compute infrastructure positions us to monetize not only AI software applications built

on our models, but also the underlying compute that powers them. As we continue to scale our terrestrial compute

infrastructure to support internal model development, training, and inference workloads, we intend to sell our high-

performance compute capacity to a limited number of third party customers.

Deploy orbital AI compute at scale. We believe growth of the projected $26.5 trillion-dollar AI market will be

constrained by Earth’s inability to rapidly scale power generation, underscoring the challenge of achieving terawatt-

scale compute without harming people and the environment. While we expect terrestrial power generation to

continue to grow, we believe the physical, environmental, and regulatory constraints will prevent it from delivering

the orders-of-magnitude increases needed to match future energy demands of the AI era. Power from the Sun, an

enormous, free fusion reactor in the sky, represents approximately 99.8% of the solar system’s energy and offers the

only truly scalable solution to terrestrial energy constraints. By combining virtually unlimited solar power in space

with our industry-leading launch costs and satellite manufacturing capabilities, we believe we can deliver compute

over time at a fundamentally lower cost structure than is possible on Earth. By the end of the decade, we intend to

deploy the first modular orbital AI compute shells and begin monetizing capacity through the sale of AI software

and AI compute. We aim to launch 100 gigawatts of AI compute capacity on solar-powered satellites each year,

equivalent to roughly one fifth of total annual U.S. power production in 2025. The amount of compute capacity we

can launch depends on three components—payload, satellite capacity, and launch frequency. With respect to

payload, Starship V3 is designed to deliver 100 metric tons to space in a fully reusable configuration while enabling

rapid turnaround times, and future generations could reach 200 metric tons, potentially as soon as Starship V4. With

respect to satellite capacity, we expect solar cells optimized for the space environment will be produced at a rapid

rate, with early satellites generating 100 kilowatts of compute power and scaling from there. Finally, with respect to

launch frequency, we expect to be able to scale to thousands of launches per year. Together, we expect these

achievements will allow us to transport approximately one million metric tons to orbit annually, powering 100

gigawatts of AI compute. Such compute capacity will also play a critical role in advancing our human augmentation

vision by expanding the reach, speed, and capability of AI beyond what is possible with terrestrial compute

infrastructure alone.

We believe we are well-positioned to execute and deliver orbital AI compute to build the infrastructure of the future.

We believe orbital AI compute is an incredibly difficult challenge that only we can solve at scale in the near term.

Design and manufacture our own chips. We plan to deepen our strategic collaboration with Tesla and Intel through

Terafab. In connection with such collaboration, we have agreed with Tesla on a general framework for the future

development of Terafab. Any specific projects undertaken pursuant to this framework will be subject to separate

negotiations and agreements (including any development timelines, milestones and capital expenditures) and have

not yet been determined. We expect Terafab to be the world’s largest chip manufacturing facility, with the goal of

170

eventually achieving one terawatt of annual compute production capacity. While Terafab is intended to expand our

internal chip manufacturing capabilities, we expect to continue sourcing a significant portion of our compute

hardware from third-party suppliers. We view Terafab as complementary to these relationships, enabling us to

augment our access to compute hardware at massive scale and further complete our highly vertically integrated

compute platform by extending our control to the foundational chip layer. By developing end-to-end capabilities

spanning the design of lithography masks, fabrication of logic and memory chips, and advanced packaging, all in a

vertically integrated closed-loop single plant, we will be able to more rapidly iterate to improve chip design and

performance. We plan to design chips that are optimized for the space environment. This collaboration directly

enables our planned orders-of-magnitude increases in AI compute deployment in orbit which would be constrained

by pure reliance on external foundries. Leveraging shared engineering resources, intellectual property, and

infrastructure across Tesla and SpaceX, as well as Intel’s expertise in designing, fabricating and packaging ultra-

high-performance chips at scale, Terafab is designed to create powerful ecosystem synergies that accelerate

innovation cycles and reduce costs. Just as we manufacture approximately 80% of Starship in-house, we expect

significant speed and cost advantages from Terafab’s vertical integration. We believe this integration, if achieved,

will provide us with a critical competitive advantage in the race to scale AI infrastructure, especially as we begin our

orbital AI compute satellite deployments.

Launch digital human augmentation. In partnership with Tesla, we are developing Macrohard, an agentic platform

designed to fully emulate digital workflows and augment human operation of computers—from coding and product

development to management and entire business processes. Similar to how autonomous systems emulate human

inputs to execute complex tasks, Macrohard is designed to augment how humans operate computers and tools to

analyze, create, and manage workflows. Unlike other enterprise software and AI applications that primarily digitize

workflows and systematize historical processes, our solutions are designed to operate as real-time, intelligence-

driven extensions of the user. Macrohard aims to combine our frontier AI model with Tesla’s physical AI prowess to

achieve the goal of augmenting the operational functions of entire companies. We expect Macrohard to benefit from

running on both state-of-the-art processors and cost efficient Tesla processors, a critical advantage of our vertical

integration. We believe Macrohard has the potential to fundamentally transform how companies across all industries

are structured and operate, thereby allowing dramatic increases in human productivity and prosperity.

Future Markets

We aim to build the infrastructure of the future in Space, leveraging our foundational competitive advantage, the

ability to launch mass at scale. By opening access to space to industries on Earth, we can grow our business by

creating new markets. Our technological capabilities enable us to repeatedly create new markets by pushing the

boundaries of what space can support. As we continue to advance and scale, we expect to unlock new market

opportunities. Over the long-term, we expect our Starship-enabled opportunities to include:

•Point-to-point terrestrial travel. We plan to develop ultra-fast long-haul point-to-point Earth transport using

Starship, enabling passengers and cargo to travel between major cities in a fraction of current transit times,

revolutionizing global logistics and passenger travel with unprecedented speed and efficiency.

•Space tourism. With meaningful advances in space technology and the continued build-out of orbital flight

infrastructure, we expect increasing interest in human space travel as it becomes easier and more common to

access space.

•In-orbit manufacturing. We aim to establish in-space manufacturing facilities that leverage the unique

microgravity conditions of space to produce materials, pharmaceuticals, and advanced components that are

difficult or impossible to manufacture on Earth, opening new high-value industrial markets.

•Passenger and cargo transport to the Moon and Mars. We intend to support large-scale passenger and cargo

missions to the Moon and Mars, delivering the people, equipment, and supplies needed to establish permanent

human settlements and accelerate the path to becoming a self-sustaining multiplanetary civilization.

•Energy production on the Moon and Mars. We aim to develop large-scale solar energy production on the

Moon and Mars, taking advantage of the thin atmosphere and constant solar exposure to generate power for

manufacturing, habitats, and future infrastructure at scale.

171

•Manufacturing capabilities on the Moon and Mars. We plan to build manufacturing infrastructure on the

Moon and Mars that utilizes local resources to produce fuel, construction materials, and other essential

resources, reducing dependence on Earth resupply and enabling sustainable long-term presence.

•Asteroid mining. We plan to pursue asteroid mining operations to extract metals and other critical resources

from near-Earth and main-belt asteroids, providing abundant raw materials for space-based industries and

reducing the need to launch mass from Earth.

Our Market Opportunity

We believe space represents the largest economic frontier in human history. Our innovations and technological

advancements are redefining existing industries and creating new market opportunities across Space, Connectivity

and AI. We believe we have a distinct ability to identify, develop, and commercialize new multi-trillion-dollar

markets that did not previously exist. We currently stand alone in our ability to deliver revolutionary breakthroughs

across spaceflight and exploration, global connectivity, and artificial intelligence, enabling an age of abundance that

we believe has the potential to propel an unprecedented expansion in the global economy.

By pioneering the world’s first and only fleet of reusable rockets at scale, we revolutionized space access through

dramatically lower cost and unmatched reliability. Lowering costs by orders of magnitude creates entirely new

industries on Earth and in space that were technologically and economically infeasible for others to access

historically. Our first trillion-dollar market was Starlink, a satellite service supported by our low-latency, high-speed

LEO constellation that required the rapid, low-cost deployment of millions of kilograms of hardware into orbit. Our

Starlink constellation powers a global connectivity platform capable of supporting broadband and mobile services,

enabling high-speed internet access to homes, enterprises, governments, and mobile users across virtually any

location on Earth. We believe our next trillion-dollar market is AI compute, which we contemplate will leverage our

rockets and satellites for massive orbital deployment.

We believe we have identified the largest TAM in human history. We estimate that our quantifiable TAM is $28.5

trillion, consisting of $370 billion in Space from space-enabled solutions; $1.6 trillion in Connectivity across $870

billion in Starlink Broadband and $740 billion in Starlink Mobile as well as additional opportunities in enterprise

and government; $26.5 trillion in AI across $2.4 trillion in AI infrastructure, $760 billion in consumer subscriptions,

$600 billion in digital advertising, and $22.7 trillion in enterprise applications. For illustrative purposes of sizing our

addressable market opportunity, we exclude China and Russia from our global estimates.

In addition to the markets we serve today, we believe we are poised to catalyze transformative breakthroughs and

create entirely new markets. Given these are longer-term opportunities at earlier stages of development, we do not

quantify them in our TAM estimates; however, we believe that over time each of these markets could eventually

represent multi-trillion-dollar economic opportunities. These new markets include long haul point-to-point terrestrial

travel, space tourism, in-orbit manufacturing, asteroid mining, energy production and manufacturing on the Moon

and Mars, and passenger and cargo transportation to the Moon and Mars.

172

SpaceX’s Estimated TAM by Segment

Space

While the size of the space market is massive for any company to address, our capabilities in space represent a

foundational competitive advantage that allow us to address markets that represent significant portions of global

gross domestic product (“GDP”)—connectivity and AI. We estimate a total market opportunity of $370 billion

across space-enabled solutions, with the lunar economy presenting a significant upside not included in the estimate.

Space-Enabled Solutions. According to Novaspace, space-enabled solutions represented a $370 billion market in

2025, including spacecraft manufacturing, launch services, satellite operations, positioning, navigation and timing

(“PNT”) devices and value-added services, as well as uncontracted costs of government space agencies. Both

commercial and government customers participate in this market, with growing space-based defense budgets

reflecting prioritization of security, resilience, and strategic autonomy by governments globally. For the purpose of

sizing our TAM, we exclude the value of satellite communications services, as we include those within our

Connectivity segment.

Lunar Economy. We believe the development of a sustained human and commercial presence on the Moon has the

potential to give rise to a new lunar economy encompassing transportation, infrastructure, communications, energy,

manufacturing (including the production of satellites and advanced chips), resource extraction, and scientific and

commercial activity. Early demand is already emerging from government space agencies and research institutions,

and we expect this to expand over time to include commercial enterprises seeking to leverage the Moon as a

platform for logistics, industrial activity, and deep-space exploration. Establishing a lunar economy requires first

proving reliable extraction of water ice to sustain life and producing hydrogen-oxygen propellant, alongside building

power, transport, and storage infrastructure in an extreme, high-cost environment. If achieved, we believe these

same resources and the Moon’s low gravity unlock the potential for scalable growth through an efficient fuel

production and refueling hub, creating a strategic access point that can potentially support deeper space

industrialization and serve as a stepping stone to establishing a civilization on Mars. Although we believe the

potential size and scope of the lunar economy is extraordinarily large, we are not providing an estimate of the TAM

for this opportunity at this time because expectations regarding the timing, pace of adoption, regulatory frameworks,

and ultimate scope of commercial activity beyond Earth are rapidly evolving alongside the development and

deployment of the technology necessary to establish a lunar presence (such as Starship). As the Moon transitions

173

from a scientific outpost into an industrial frontier, SpaceX is positioned to spearhead this revolutionary expansion,

and we believe that continued advancements in our launch capabilities, space infrastructure capabilities, and cost

efficiency will allow us to meaningfully accelerate the development of a sustainable lunar economy.

Connectivity

We believe the global connectivity market represents a substantial and durable opportunity, driven by the increasing

reliance of consumers, enterprises, and governments on high-speed, low-latency, reliable connectivity across both

terrestrial and remote environments. Across Starlink Broadband and Starlink Mobile, we estimate a total market

opportunity of $1.6 trillion reflecting primarily consumer use cases. We believe these traditional use cases, however,

do not account for the long-term market opportunity, as connectivity is evolving into a critical infrastructure layer

underpinning the global economy, enabling entirely new categories of demand. As high-performance, ubiquitous

connectivity becomes embedded across transportation networks, autonomous systems, and smart devices, we expect

the scope of the market to extend well beyond the traditional definitions.

Starlink Broadband. The global demand for ubiquitous, high-speed broadband internet creates an approximately

$870 billion dollar opportunity. Our satellite broadband service, Starlink, is positioned to capture value across

multiple massive and rapidly expanding markets:

•Consumer Broadband. As the digital economy continues to expand, ubiquitous, high-speed, reliable internet

has become a structural necessity for households worldwide—powering opportunity and the next wave of

global prosperity. According to Euromonitor, there were approximately 1.8 billion global households in 2025.

As Starlink develops, we believe that our broadband network can connect, and improve the existing connection,

of every household globally. Given varying economic conditions and consumer purchasing power across

different countries, we use a different monthly ARPU for different parts of the world based on country-specific

consumer broadband ARPU from Omdia as we seek to make our service affordable and accessible across

different economic development contexts. Region-specific ARPU assumptions result in a weighted average of

$31 monthly ARPU for residential broadband internet services globally, according to Omdia. This global

average consists of a weighted average monthly ARPU of $43 in high-income markets, $16 in upper-middle

income markets, and $9 in lower-middle income and low income markets per World Bank classification.

Together this represents a total addressable market of $660 billion based on 1.8 billion households.

Approximately 40% of the global population lives in rural areas, remaining structurally underserved by

terrestrial broadband infrastructure due to unfavorable deployment economics, limited network density and high

last-mile costs.

This structural imbalance creates a large, durable and relatively uncontested baseline market for satellite-based

connectivity solutions. For many of these households, Starlink represents the first viable option for high-speed,

low-latency internet access, with limited competition from terrestrial providers. Unlike terrestrial networks,

which require significant incremental capital to extend coverage to low-density areas, our space-based

architecture enables economically scalable service delivery across these regions with minimal marginal cost per

additional user.

Importantly, while rural and underserved geographies provide a compelling initial adoption vector, we believe

Starlink’s value proposition extends well beyond these markets. As network capacity increases and product

performance continues to improve, we expect to compete increasingly in suburban and urban environments.

Accordingly, while rural households represent a large and durable entry point for our connectivity offering, we

view this segment as a foundational layer upon which significantly broader consumer, enterprise and

government demand can be built.

•Enterprise Solutions.We offer fixed site broadband solutions tailored for the needs of our enterprise customers

across many different industries, including construction, agriculture, retail, telecom, hospitality and others. For

the purpose of sizing market opportunity, we include small and medium sized businesses within our Enterprise

Solutions market opportunity. Our Starlink enterprise offerings can provide important primary or back-up

connectivity for every business in the geographies where we are licensed to operate. According to Grand View

174

Research, the global business broadband market in 2025 across small to medium sized business and enterprise

usage is estimated to be $200 billion.

•Government Solutions. Driven by increasing demand for resilient, low-latency, and highly secure

communications in contested and remote environments, defense organizations and governments around the

world are increasingly turning to commercial satellite providers with connectivity solutions to supplement and

enhance traditional military networks. According to Novaspace, the global satellite communications market

driven by defense and government demand in 2025 was $5 billion. The estimate of the government

communications market includes only publicly disclosed programs and budgets and does not include classified

missions or other restricted uses, which we believe represent additional sources of demand.

Starlink Mobile. According to Omdia, as of December 31, 2025, there were eight billion mobile connected devices

globally. We believe our Starlink Mobile offering will be able to provide continuous global coverage and

substantially reduce mobile “dead zones,” which remain areas that are structurally underserved by the limitations of

the networks of current mobile network operators. For example, according to the J.D. Power U.S. Wireless Network

Quality Performance Study, U.S. wireless customers experienced service problems in approximately one out of

every 11 mobile interactions, even in well-connected areas. In addition, an estimated 40% of the global population

resided in rural areas in 2024 according to the World Bank, where terrestrial mobile coverage can be limited or

unreliable. While we expect Starlink Mobile service today to be most impactful for customers in remote areas

uncovered by terrestrial mobile networks, as our constellation grows and our product performance continues to

improve, we will compete to be the preferred connectivity experience to our customers no matter where they are

located, whether in rural, suburban, or urban areas. The next-generation of Starlink Mobile satellites, in combination

with our recent purchase of wireless spectrum from EchoStar, is designed to provide high bandwidth and low

latency connectivity directly to end user devices, enabling a connectivity solution on par with terrestrial mobile

networks. Given varying economic conditions and consumer purchasing power across different countries, we

assume a different monthly ARPU for different parts of the world as we seek to make our service affordable and

accessible across different economic development contexts. Our region-specific ARPU assumptions result in a

weighted average monthly mobile ARPU of $8 per user. This global average consists of a weighted average monthly

ARPU of $18 in high-income markets, $5 in upper-middle income markets, $2 in lower-middle, and $2 in low

income markets. Based on the total number of connected devices globally and the mobile ARPU, we estimate the

Starlink Mobile market opportunity to be $740 billion. We expect to continue to partner with mobile network

operators globally as we expand coverage and participate in the broader mobile connectivity market.

Additional and Future Starlink Applications. We believe the long-term market opportunity for Starlink extends

materially beyond traditional fixed broadband and satellite-to-mobile connectivity. Many of these use cases

represent new categories of demand that were not previously addressable with legacy terrestrial or satellite solutions

due to limitations in coverage, latency, capacity, or cost. While these additional and future use cases are early stage

and not yet captured in conventional industry market definitions, we believe they have the potential to significantly

expand the total addressable market for connectivity over time.

•Enterprise Mobility. Because our Starlink solutions are uniquely well-suited for in-motion environments,

remote, or hard-to-serve locations, we are able to provide high-performance connectivity across land, air, and

sea. We believe we have a differentiated right to win these verticals as existing connectivity solutions are not

able to provide sufficient speed, latency and reliability, with frequent service outages driven by weather, orbital

mechanics and coverage gaps. Our Starlink constellation directly addresses these deficiencies, creating a

compelling path for us to capture a substantial share of opportunities and to unlock previously unattainable

levels of service quality and customer willingness to pay.

In land mobility, Starlink supports connectivity for vehicle fleets, including trucking, rail, public safety vehicles,

and autonomous systems, enabling real-time telematics, route optimization, safety monitoring, and onboard

passenger connectivity, as fleets become increasingly connected and data-driven.

In aviation, Starlink delivers high-speed, low-latency in-flight connectivity for commercial airlines, business

aviation, and government aircraft, supporting passenger broadband, operational communications, and real-time

aircraft data transmission, as airlines increasingly prioritize differentiated onboard experiences and operational

175

efficiency. There are approximately 23,900 commercial aircraft, according to Oliver Wyman, and

approximately 24,500 privately owned aircraft, according to Corporate Jet Investor, in the world, which can be

served by our aviation offering.

In maritime, Starlink provides connectivity for commercial shipping, offshore energy platforms, cruise lines,

and government vessels, enabling crew welfare, operational optimization, safety systems, and real-time data

transfer, as connectivity becomes a standard requirement across global fleets. Our potential customer base as of

2025 consists of approximately 99,000 commercial merchant ships, defined as being 100 gross tons or more,

approximately 21,000 fishing vessels, and approximately 4,000 cruise ships and private yachts, according to

Marine Traffic Dashboard.

•Expanded Enterprise and Government Applications

Enterprise Back-Up and Failover Connectivity. As connectivity becomes a mission critical component of

enterprise operations, we believe back-up and failover connectivity is evolving into a foundational layer of

enterprise infrastructure. The increasing cost of downtime, combined with the proliferation of cloud-based and

latency-sensitive applications, is driving enterprises to prioritize uptime, business continuity, and network

resilience and adopt multi-layered connectivity architectures. We believe this shift will result in a meaningful

expansion of the connectivity market.

Expanded Government Applications. We believe traditional connectivity market estimates do not fully

capture the scope of government-related demand, particularly in mission-critical and classified applications. The

growing importance of secure communications, real-time intelligence, and resilient network architectures is

driving sustained investment in connectivity capabilities across defense and civilian agencies. These use cases

tend to command higher value and longer-duration contracts, contributing to a meaningful and durable

expansion of the connectivity market.

Smart Device Connectivity. The proliferation of connected devices across various physical environments—

including sensors, wearables, vehicles, appliances, and infrastructure systems—is driving increasing demand for

ubiquitous, reliable, and low-latency connectivity. As of 2025, there were approximately 22 billion IoT

connected devices globally, forecasted to reach 47 billion by 2031. As billions of connected devices generate,

transmit, and act on data, connectivity becomes an essential enabler of new categories of economic activity. As

these devices grow in scale into the tens of billions globally and become more intelligent and data-intensive, we

believe the scope of the connectivity market will expand significantly beyond traditional human-centric usage.

In-Orbit Data Transport. We operate a large constellation of over 23,000 inter-satellite lasers that create a

dynamic mesh network in space and enable traffic rerouting through orbit. We believe this laser mesh network

will help us unlock a new connectivity market by enabling third-party satellites to utilize our in-orbit data

transport layer. While most of our laser mesh network capacity is used to power our Starlink services, we

selectively monetize excess capacity through our Plaser program. We allow third parties to purchase our space

laser hardware and connect their satellites to our Starlink network, allowing them to offload data to a ground

station anywhere on Earth while bypassing the need to build their own relay architecture or ground stations. As

satellite constellations grow, we expect market demand for high-throughput, low-latency data relay to increase

across commercial and government operators. While this market remains nascent, we believe the opportunity

represents a meaningful expansion beyond traditional satellite connectivity TAM.

Artificial Intelligence

The market for artificial intelligence is currently undergoing explosive structural growth, emerging as a foundational

utility for the modern global economy and unlocking a multi-trillion-dollar opportunity. Our frontier models,

consumer and enterprise applications, and AI infrastructure solutions are strategically positioned to capture value

across four key components of this vast ecosystem, resulting in an estimated total market opportunity of $26.5

trillion.

AI Infrastructure.According to RAND Corporation, global data center compute demand is estimated to be 235

gigawatts in 2030, of which 70% is estimated to be utilized for AI workloads. Assuming a target Power Usage

176

Effectiveness of 1.2 and an all-in chip power consumption per GPU of 1.3 kilowatts per GPU—that of an H100

SXM—this AI workload demand corresponds to 104 million GPUs required. We apply an 80% utilization rate per

the National Electrical Installation Standards and a GPU rental rate of $3.33 per hour, according to Silicon Data,

which is based on the median of neocloud GPU rental rates in 2025; we note that the rental rate has historically

varied subject to market conditions. As a result, we estimate the AI compute infrastructure market opportunity to be

approximately $2.4 trillion.

Consumer Subscriptions.As demand for AI solutions surges, fueled by widespread adoption of AI tools that

enhance productivity, creativity, personalization, and real-time assistance in everyday life, consumers are

increasingly turning to subscription-based access to high-performance AI platforms. These platforms, equipped with

advanced reasoning, seamless real-time data integration, and multimodal capabilities, are essential in today’s ever-

more receptive and interconnected world. We believe SpaceX is well positioned to address this opportunity through

our X and Grok platforms by delivering a differentiated product centered on truth-seeking and real-time relevance.

Our roadmap for future models contains multi-trillion parameter models, which could represent a step change in

reasoning depth and overall intelligence. Through Grok’s integration with X and proprietary access to real-time data

inflows, we believe we can better address a broader set of high-frequency, high-value consumer use cases and

increase user engagement and willingness to pay, positioning Grok to capture a larger share of the consumer AI

subscription market relative to standalone, non-integrated offerings. We estimate our market opportunity based on

the global population of individuals aged 10 and over in 2025—approximately five and a half billion according to

Euromonitor—multiplied by the weighted average monthly subscription revenue of $12, resulting in an annualized

market opportunity of approximately $760 billion. Our weighted average monthly revenue assumes different

monthly subscription fees across different geographies around the world. We assume $30 monthly cost of a

SuperGrok subscription in high-income countries, $8 monthly cost in upper-middle and lower-middle income

countries, and significantly lower monthly cost in low income countries, as defined by the World Bank.

Digital Advertising.Digital advertising represents a large and growing global market opportunity as businesses

increase marketing budgets towards digital platforms that enable targeted advertising, measurable performance, and

direct engagement with consumers. In 2025, global digital advertising spending totaled $600 billion according to

S&P Global Market Intelligence. We believe that X’s ability to combine large-scale user engagement, real-time

content, and advanced AI-driven performance marketing tools positions us well to participate in this significant

market opportunity.

Enterprise Applications. AI is revolutionizing enterprise applications as organizations across industries increasingly

adopt AI solutions to automate complex workflows, augment knowledge workers, enhance decision-making,

redefine productivity, and improve operational efficiency. Specifically, we believe that our enterprise applications,

including Macrohard, agentic AI, will increasingly support knowledge workers across industries by automating

routine cognitive tasks, assisting with research and analysis, generating content and code, and refining decision-

making processes. Ultimately, we believe this transformation could evolve knowledge workers into empowered

managers of autonomous agents, unlocking unprecedented levels of creativity and productivity.

We believe we are still in the early days of AI transforming enterprises, with AI-powered enterprise applications

poised to reshape the digital economy. The Digital Cooperation Organization (“DCO”) defines the digital economy

as economic activity reliant on, significantly enhanced, or enabled by digital technologies and their applications,

including the following products and services: AI and advanced analytics, blockchain and decentralized

technologies, cloud services, digital connectivity, digital devices and the IoT, encryption and cybersecurity,

immersive technologies, and robotics and autonomous systems. DCO estimates that the digital economy will grow

three times faster in 2026 on a year-over-year basis compared to the estimated growth of the global GDP, reaching

approximately $22.7 trillion in 2026. In a survey of CTOs, senior technologists, policymakers, and digital economy

experts, also conducted by DCO, AI and advanced analytics were identified by 69% of respondents as their top

digital technology priority—higher than any other surveyed priority. We believe that our enterprise strategy, which

is focused on serving the digital needs of the world’s largest industries with AI solutions, positions us competitively

to pursue this rapidly growing opportunity.

177

Future Markets

Beyond the established markets reflected in our TAM, we envision that ongoing advancements in our technology

and infrastructure will unlock entirely new markets over time. As launch costs decline, satellite capabilities advance,

and large-scale compute infrastructure expands, innovative applications and new markets may emerge that harness

our integrated infrastructure across space, connectivity, and AI. Although these prospects remain nascent, with

uncertain timing and scale—and thus are excluded from our quantified total addressable market estimates—we

believe they hold trillions of dollars of eventual potential for groundbreaking innovation and value creation,

eventually representing multi-trillion-dollar economic opportunities.

Long-Haul Point-to-Point Terrestrial Travel. Our Starship vehicle has the potential to revolutionize terrestrial

commercial transportation by achieving an unparalleled combination of speed, reliability and cost efficiency. This

capability could reduce most international long-haul flights to under 30 minutes, enabling point-to-point travel to the

furthest location in an hour or less. While we must surmount technological, economic and regulatory obstacles to

fully capitalize on this opportunity—such as restrictions on supersonic flights over land in certain regions due to

sonic booms, and the economic feasibility of shorter routes—we believe we are strategically positioned to take share

of the terrestrial logistics and transportation market.

Space Tourism. Historically, human spaceflight has been limited to government astronauts, augmented by a limited

number of privately funded missions. Yet, with meaningful advances in space technology and the ongoing

expansion of orbital flight infrastructure, we anticipate a gradual increase in accessibility of spaceflight over time,

potentially enabling a new category of commercial human spaceflight and tourism. Under 30 people out of the

global population visited Earth’s orbit in 2025, which we believe could be a far greater number in the future.

Passenger and Cargo Transport to the Moon and Mars. Looking further ahead, advances in reusable launch

systems and deep-space transportation infrastructure may enable new forms of interplanetary logistics, including

passenger and cargo transportation to the Moon and Mars. Supporting a sustained human presence on another planet

would require the regular transport of people, equipment, and materials at a scale not previously possible.

Energy Production and Manufacturing on the Moon and Mars.Establishing a sustained human and industrial

presence on the Moon and Mars would require reliable, large-scale energy generation to support habitats,

manufacturing, and scientific operations. Potential solutions could include solar power systems, taking advantage of

the thin atmosphere, constant solar exposure, and other advanced energy technologies designed to operate in the

unique environmental conditions of the Moon and Mars. Over time, we believe that advances in planetary

infrastructure may enable manufacturing on the Moon and Mars using locally available resources.

In-Orbit Manufacturing. Terrestrial manufacturing is inherently constrained by gravity, which imposes

fundamental limitations on processes at the atomic and molecular level. Establishing in-orbit infrastructure unlocks

large-scale, high-value production free from those traditional barriers, enabling breakthroughs in precision and

efficiency. The microgravity environment of space fosters innovative advancements in key industries, such as

pharmaceuticals—where it enhances drug solubility, purity, crystallization, and stability—as well as, advanced

materials and semiconductors, allowing for superior crystal formation and material properties unattainable on Earth.

Beyond these particle-level innovations, in-orbit facilities overcome Earth’s energy constraints by harnessing

abundant, uninterrupted solar power, facilitating energy-intensive operations with unparalleled sustainability.

Asteroid Mining.Asteroid resources, including platinum-group metals, rare earth elements, nickel, cobalt, iron and

water, represent a vast untapped reservoir beyond Earth’s gravity well, with some near-Earth objects containing

concentrations of elements far exceeding typical terrestrial ore grades. With meaningful advances in reusable launch

capabilities, autonomous robotics, and in-situ processing technologies, we believe the accessibility of asteroid

resources will expand over time, unlocking a new category of commercial space resource extraction. We believe our

experience in launch systems, spacecraft development, and space infrastructure uniquely positions us to pursue

asteroid mining operations to extract metals and other critical resources from near-Earth and main-belt asteroids,

providing abundant raw materials for space-based infrastructure, reducing the need to launch all mass from Earth.

178

Our Solutions & Services

Unparalleled Launch Capability

Our unmatched launch capability is the foundational competitive advantage that enables our unique solutions and

services. We are the market leader in orbital launch, providing low-cost, reliable, and frequent access to space for

commercial and government customers. Our launch services are built around a fleet of reusable rockets and

spacecraft. SpaceX’s family of rocket systems and spacecraft address missions ranging from routine cargo delivery

to the International Space Station to deep-space exploration. The Falcon class of rockets delivered over 80% of mass

to orbit in the year ending December 31, 2025. Starship, a two-stage super heavy-lift launch vehicle that we have

been flight testing since 2023, further enhances our industry-defining launch offerings.

Separate from our fleet of reusable rockets, SpaceX’s launch advantage is equally underpinned by our fleet of

advanced spacecraft. Our International Space Station cargo and human spaceflight missions are launched on Falcon

9 and flown on the Dragon crew and cargo spacecraft. The vehicles autonomously dock to the station, delivering

pressurized and unpressurized cargo, and passengers. Both Dragon variants are partially reusable and perform fully

autonomous rendezvous, docking, and return operations.

Our Fleet of Launch Vehicles and Spacecraft

Our Fleet of Launch Vehicles

Falcon 9. The Falcon 9 rocket is a reusable, two-stage rocket designed and manufactured by SpaceX for the safe,

reliable, and cost-effective transport of satellites, scientific payloads, cargo, and crew to Earth orbit and beyond.

Powered by liquid oxygen and rocket-grade kerosene, the first-stage is equipped with nine Merlin 1D engines

producing over 1.7 million pounds of thrust at sea level, while the second stage utilizes a single vacuum-optimized

Merlin engine for precise orbital insertion. First launched in 2010, Falcon 9 is the world’s first orbital-class rapidly

reusable rocket, and has become the most active orbital launch vehicle today, with approximately 620 orbital space

launches as of March 31, 2026 and an over 99% mission success rate. Falcon 9 is capable of delivering

approximately 23 metric tons to LEO and eight metric tons to geosynchronous transfer orbit. Reusability allows

SpaceX to refly the most expensive parts of the rocket, which in turn drives down the cost of space access. Falcon

9’s reusable components primarily include its booster, which lands on one of our autonomous drone ships out on the

179

ocean or on one of our landing zones near our launch pads ahead of being refurbished for a future launch, and its

payload fairing halves, which are recovered via parachute-assisted splashdowns and are refurbished and reused after

retrieval. The second stage is not designed for recovery or reuse and instead safely deorbits after successful payload

deployment.

Falcon 9 Overview

Falcon 9 introduced a combination of technical innovation, cost reduction, and operational scale that materially

altered the economics of orbital launch and established our position as the leading commercial launch provider.

•First Orbital-Class Rapidly Reusable Rocket: In December 2015, Falcon 9 achieved the first vertical landing

of an orbital-class booster, followed in April 2016 by the first autonomous drone ship landing in the Atlantic

Ocean. Reuse of boosters and fairings, a practice pioneered by SpaceX in the launch industry, fundamentally

enables our launch rate and capacity and forms the basis for the launch system’s inherent reliability. Through

recovering, inspecting, and evaluating flown hardware, SpaceX gains insight into system performance that

would not be otherwise achievable. Partial reusability for orbital spaceflight has reduced cost per ton to orbit by

approximately 85% as compared to the historical average launch cost per kilogram of $18,500.

•Reusability Enabled Cost Structure Advantage: Reuse of the first-stage—representing the majority of

vehicle manufacturing cost—has materially reduced marginal launch costs relative to fully expendable systems.

•Highest Operational Tempo in History: With approximately 620 orbital space launches over 15 years of

operation, Falcon 9 is the most frequently flown active orbital launch vehicle to date. In 2025, Falcon 9

conducted 165 launches, accounting for over half of all global orbital launches in the year while delivering over

80% of mass to orbit.

•Track Record of Success: As of March 31, 2026, Falcon 9 has achieved an over 99% mission success rate.

Falcon 9 has achieved over 530 successful booster landings and more than 540 launches completed by a flight-

proven Falcon rocket, underscoring the reliability of its reusability architecture.

•Human Spaceflight Certified: Falcon 9, paired with SpaceX’s Dragon crew spacecraft, is the only U.S.-based

launch vehicle certified by NASA under the Commercial Crew Program to transport astronauts to and from the

180

International Space Station. As of December 31, 2025, Falcon 9 has successfully launched 19 human

spaceflight missions with a 100% mission success rate.

•In-House Engine Development and Manufacturing: Falcon 9 is powered by Merlin engines that are

designed, developed, and manufactured in‑house, providing vertical integration across propulsion design,

production, and testing. The Merlin engine achieves one of the highest thrust‑to‑weight ratios of any rocket

engine in operational service, contributing to Falcon 9’s performance and payload capacity.

Falcon 9

As we transition primary production and development resources toward the fully and rapidly reusable Starship

system, Falcon 9 continues to serve as the backbone of our launch revenue base; generating high-margin recurring

cash flows while providing critical operational experience in high-cadence reuse. The proven capabilities of Falcon

9 established us as the leading provider of launch services globally and laid the technological and economic

foundation for the next era of space transportation.

Falcon Heavy. Falcon Heavy is a partially reusable super heavy-lift launch vehicle, designed to deliver large

payloads to orbit. Building on the proven architecture of the Falcon 9 rocket, Falcon Heavy is composed of three

reusable Falcon 9 nine-engine boosters whose combined 27 Merlin engines generate more than five million pounds

of thrust at liftoff—one of the most powerful operational rockets in the world today. It is capable of carrying

approximately 64 metric tons of payload to LEO and 27 metric tons to geosynchronous transfer orbit. Falcon

Heavy’s reusable components primarily include its three boosters, which are designed to land vertically on drone

ships in the ocean and landing zones near our launch sites, and its payload-faring halves, which are recovered via

parachute-assisted splashdown and are refurbished and reused after retrieval. The second stage is not designed for

recovery or reuse and is designed to safely deorbit after successful payload deployment, similar to Falcon 9.

181

Falcon Heavy Overview

•Reusability: Falcon Heavy incorporates a design focused on reusability, which has contributed to lowering the

cost of access to space and altering the launch industry’s economic model for large or high-value payloads. The

vehicle’s two side boosters, equipped with hypersonic grid fins and advanced propulsion systems, enable

controlled recovery and soft landings. This capability enables reusability, with missions launching on flight-

proven boosters generally priced below those of traditional expendable flights. Our Falcon 9 boosters, which are

qualified for up to 40 flights, are also used on Falcon Heavy, with an average of 6 flights per booster on Falcon

Heavy. Although our Falcon 9 boosters have been engineered and demonstrated to support up to 40 flights, we

have established a maximum accounting useful life of 25 flights as an estimate based on forecasted utilization.

This estimate reflects: (i) our strategic transition to Starship, which is expected to materially reduce future

Falcon 9 flight demand; and (ii) restrictions under certain government contracts that prohibit the use of boosters

flown more than five times on their missions. These useful life estimates are periodically reassessed based on

engineering qualification data, post-flight inspections, recovery success rates, actual fleet performance, cost

sensitivity analyses, and the long-range launch manifest.

182

Falcon Heavy

•Exploratory Missions Beyond Earth’s Orbit: Falcon Heavy first launched in February 2018, when it put a

Tesla Roadster and its mannequin passenger, Starman, into orbit around the Sun. This was the first instance of a

car sent into deep space and demonstrated the rocket’s capability for trans-Mars injection. Since its inaugural

flight, Falcon Heavy has completed missions that expanded the scope of space exploration and commercial

spaceflight. Falcon Heavy has been selected by NASA to launch critical weather satellites, interplanetary probes

including Europa Clipper (Jupiter) and Dragonfly (Saturn), and the upcoming Nancy Grace Roman telescope,

designed to study exoplanets and dark energy and matter.

183

Starman in Orbit

•Perfect Performance Record: As of March 31, 2026, Falcon Heavy had successfully completed 11 launches,

all resulting in successful payload delivery. Falcon Heavy flown boosters have also safely completed 18 total

recoveries and 16 reflights. It was certified for National Security Space Launch in 2019, authorizing its use for

U.S. government operations alongside Falcon 9.

Starship. A fully reusable two-stage super heavy-lift launch vehicle, Starship stands to fundamentally transform

spaceflight by making it more accessible, cost-effective, and scalable than ever before. Comprising the Super Heavy

booster (powered by 33 Raptor engines) and the Starship upper stage (with three sea-level and three vacuum Raptor

engines), Starship V3 is designed to deliver 100 metric tons to space in a fully reusable configuration while enabling

rapid turnaround times akin to commercial aviation, and future generations could reach 200 metric tons, potentially

as soon as Starship V4. To date, we have executed 11 Starship flight tests. We have also scheduled a 12th flight test,

which will debut the next generation Starship vehicle and Super Heavy booster, powered by the next evolution of

our Raptor engine and launching from a newly designed pad at Starbase. We expect Starship to commence payload

delivery to orbit in the second half of 2026. We have achieved innovative milestones, including multiple successful

ascents of the world’s most powerful rocket; the launch, return, catch, and reuse of the Super Heavy booster; the

return of its upper stage within three meters of its intended landing point; the transfer of approximately five metric

tons of cryogenic propellant between tanks while in space, a first of its kind operation that provides key data for

future full-scale propellant transfer operations; successful in-space relights of the Raptor engines; and multiple

controlled reentries through Earth’s atmosphere. The purpose of flight tests is to collect data so no result, even loss

of a vehicle, is considered a failure because we learn something.

Starship is a key enabler of our growth objectives, including the deployment of next-generation V3 satellites, direct-

to-cell constellations, and orbital AI compute at scale. Achieving our targeted launch cadence with Starship will

require significant progress on several key milestones and the investment of significant capital resources. These

include: securing additional land and developing high-rate launch sites and supporting infrastructure across multiple

locations; scaling production of Starship vehicles and Raptor engines; constructing propellant production facilities,

including air separation units and methane liquefaction plants co-located with launch sites; securing sufficient power

supply; and obtaining the necessary regulatory approvals, particularly from the FAA, to support a high launch

cadence while addressing public safety and environmental considerations. Our development of Starship and its

184

associated infrastructure assumes continued successful iteration through flight testing, regulatory progress, supply

chain scaling, and cost reduction driven by increasing reusability. We have made substantial investments in

manufacturing scale-up, including Starfactory for high-volume vehicle production, multiple large-scale vertical

integration and refurbishment facilities, additional launch towers, test infrastructure, propellant production assets,

and power generation capabilities.

Full reusability of Starship’s upper stage is not required to deploy our V3 satellites and V2 Mobile satellites in low-

Earth orbit. In-orbit refueling is also not required for any of these LEO programs and is instead intended for

missions beyond LEO, such as lunar and interplanetary transport. Starship’s substantial payload capacity to LEO,

even in partially reusable or expendable configurations, enables meaningful progress toward these objectives. We

have already demonstrated Super Heavy booster reusability in multiple integrated flight tests. As a result,

meaningful advancement across the deployment of next-generation V3 satellites, direct-to-cell constellations, and

the orbital AI compute program is not dependent on achieving full reusability.

Starship Overview

•Full and Rapid Reusability and Drastically Reduced Launch Costs: Starship’s core design innovation is its

full and rapid approach to reusability: both stages return to Earth for catch and rapid refurbishment. The Super

Heavy booster returns to the launch site following stage separation and is caught mid-air by the launch tower’s

mechanical arms, also known as “chopsticks,” to facilitate immediate inspection, refurbishment, and relaunch.

185

“Chopstick” Super Heavy Booster Catch

The Starship upper stage, after orbital delivery or missions beyond, is designed to reenter protected by advanced

heat shield tiles, execute a propulsive landing burn, and be similarly caught mid-air by the launch tower’s

mechanical arms. We believe that Starship’s full and rapid reusability will enable sub-one hour reflights,

causing a paradigm shift in launch cadence.

Starship Landing Burn

186

•Improvements in Engine Development Underpin Starship’s Massive Payload Capacity: With a payload

bay volume rivaling the pressurized sections of the International Space Station, Starship is designed to deploy

structures like space station modules, large telescopes, our next-generation V3 satellites, and future AI compute

satellites. Starship is powered by 39 Raptor engines, which are full-flow staged combustion cycle rocket engines

burning cryogenic liquid methane and liquid oxygen. Raptor engines offer nearly triple the thrust per engine,

higher efficiency, and better performance for heavy-lift and deep space missions compared to the Merlin engine

used on Falcon 9. Each Raptor 3 engine in Starship saves nearly a ton of vehicle mass compared to previous

generations by removing heat shields and simplifying plumbing. Starship’s capacity enables the next leg of our

growth, including scaling our Starlink Mobile constellation and orbital AI compute.

•Orbital Refueling: Starship’s expected orbital refueling capability will allow tanker variants to refill the upper

stage in LEO and extend its range for deep-space missions beyond Earth’s orbit. These capabilities are expected

to revolutionize mission architecture, with each Starship designed to be capable of transporting large numbers

of people or hundreds of metric tons of cargo to destinations like the surface of the Moon and Mars.

•Sustainable Human Exploration Beyond Earth: Starship was designed from the beginning to fly to other

worlds and enable self-growing bases on the Moon, an entire civilization on Mars, and ultimately expansion

beyond our solar system. As NASA’s Human Landing System for Artemis, Starship is built to deliver

astronauts and cargo to the lunar surface and serve as the key enabler for supporting permanent presence on the

Moon.

•Versatility Across Mission Profiles: Beyond deep space, Starship is designed to adapt to diverse roles

including the U.S. Space Force’s Rocket Cargo program for rapid point-to-point global logistics, Starlink and

other commercial satellite constellations, in-orbit manufacturing components and hardware, space tourism, and

others.

Starship is designed to enable a step-function advancement in our capabilities, featuring rapid, full reusability of

both the Super Heavy booster and the Starship spacecraft to achieve unprecedented throughput at significantly

reduced costs compared to existing systems. As Starship progresses toward full operational utilization, the Falcon 9

and Falcon Heavy platforms will remain key assets for specialized missions, including NASA crew rotations and

national security payloads.

Dragon Cargo Spacecraft. The Dragon cargo spacecraft is an uncrewed vehicle designed primarily for transporting

cargo to and from the International Space Station under NASA’s Commercial Resupply Services program. As an

evolution of the original Dragon spacecraft, this vehicle represents a critical component of our portfolio, enabling

reliable, cost-effective logistics for space missions. The spacecraft consists of a pressurized section for

environmentally controlled cargo and an unpressurized trunk section for additional payloads. It has a launch payload

mass of up to 6,000 kilograms and a return payload mass of 3,000 kilograms, making it uniquely suited for both

delivery and retrieval of scientific experiments, supplies, and hardware, and establishing SpaceX as the only

company capable of returning significant amounts of cargo from the International Space Station back to Earth.

187

Dragon Cargo Overview

•Key features: Key highlights of the Dragon cargo spacecraft include its propulsion system with 16 Draco

thrusters for precise orbital maneuvering, autonomous docking capabilities via NASA’s International Docking

System Standard (IDSS), and a trunk equipped with solar panels for power generation during flight.

•Launch vehicle, return mechanism and mission profiles: The spacecraft is launched atop the Falcon 9 rocket

and returns to Earth via parachute-assisted splashdown in the ocean, where it is recovered for refurbishment and

reuse. Dragon supports extended in-orbit durations, typically spending several weeks docked to the International

Space Station before undocking with returned cargo.

•Historic accomplishments: Launched by Falcon 9 in 2012, our Dragon spacecraft became the first commercial

spacecraft to deliver cargo to and from the International Space Station and, eight years later, the first privately

built vehicle to fly humans to the orbiting laboratory. This achievement ended U.S. reliance on foreign vehicles

for International Space Station resupply following the Space Shuttle’s retirement in 2011. The original Dragon

variant (later known as Dragon 1) established a critical role in advancing research on the space station as the

only spacecraft capable of returning significant amounts of cargo to Earth. The upgraded cargo spacecraft

pioneered autonomous docking without robotic arm assistance, delivered major hardware upgrades for the

station including new solar arrays, and recently debuted the ability to reboost the station’s altitude. It remains

the only reusable cargo spacecraft in operation. As of March 31, 2026, our Dragon spacecraft has completed

over 30 cargo missions to the International Space Station.

Dragon Crew Spacecraft. Dragon is engineered to fly humans to and from Earth orbit, including the International

Space Station. The spacecraft is designed to accommodate up to seven passengers, with a pressurized cabin for crew

habitation, life support systems, and cargo.

188

Dragon Orbiting Earth's Poles

•Key features: Dragon Crew spacecraft is equipped with advanced avionics, touchscreen interfaces for manual

control, and an integrated trunk with solar power generation. Dragon Crew’s propulsion includes 16 Draco

thrusters for orbital adjustments and 8 Super Draco engines for its launch escape system, enabling rapid

separation from the rocket in the unlikely event of an emergency.

•Launch vehicle, return mechanism and mission profiles: The spacecraft is launched atop the Falcon 9 rocket

and returns to Earth via parachute-assisted splashdown in the ocean, where it is recovered for refurbishment and

reuse. The design emphasizes reusability, with vehicles certified for multiple flights after refurbishment, and

supports missions lasting up to nine months on the International Space Station.

•Historic accomplishments: Revolutionary accomplishments of Dragon include being the first privately

developed spacecraft to transport humans to and from the International Space Station, achieved during the

Demo-2 mission in May 2020 which carried NASA astronauts Doug Hurley and Bob Behnken. This milestone

returned human spaceflight capabilities to the United States for the first time since the Space Shuttle’s

retirement in 2011, reducing dependence on foreign spacecraft. Dragon has enabled regular astronaut rotations

under NASA’s Commercial Crew Program, flying nearly 15 successful Crew and Private Astronaut missions to

the International Space Station to date, while pioneering space tourism by carrying commercial astronauts on

private flights. Its autonomous docking technology, life support for extended durations, and abort system have

set new safety standards achieving a flawless record in crewed operations.

Connectivity

Starlink Consumer Broadband

Starlink Consumer Broadband is a broadband network powered by our global LEO satellite constellation, designed

to deliver high-speed, low-latency internet connectivity anywhere on Earth. The service provides fiber-like

download speeds with latency low enough to support intensive real-time applications, such as content streaming,

video calls, and online gaming, while requiring only visible sight to the sky and electricity for installation. Since

launch, Starlink has scaled rapidly, serving approximately 10.3 million subscribers across 164 countries, territories,

and other markets as of March 31, 2026.

189

Starlink Consumer Broadband is enabled by the largest satellite constellation in human history with approximately

9,000 broadband satellites as of March 31, 2026, operating in LEO to deliver latency comparable to many terrestrial

broadband connections. We launched approximately 3,100 Starlink broadband and mobile satellites in 2025, which

is approximately five times more than the total number of active satellites in the entire second largest LEO satellite

constellation. We provide download speeds exceeding 400 Mbps with round-trip latencies as low as 21 milliseconds

—performance that rivals or surpasses traditional terrestrial broadband while also reaching locations no traditional

fiber or cellular network can economically serve. Satellite-based communications are uniquely suited to reach

underserved and remote areas by delivering coverage directly from LEO without requiring local infrastructure. In

contrast, terrestrial networks depend on costly, ground-based buildouts that are often uneconomical in low-density or

hard-to-access regions. As of March 31, 2026, the constellation incorporated over 23,000 inter-satellite lasers that

create a dynamic mesh network in space, enabling traffic to route through orbit rather than relying solely on

terrestrial backhaul infrastructure. Satellites autonomously maneuver to avoid collisions and are designed for

controlled end-of-life deorbit, supporting long-term orbital sustainability. Successive generations of our broadband

satellites, including V3 satellites, are expected to increase throughput, power capacity, and network efficiency, with

production vertically integrated and performed largely in-house. Our focus on vertical integration has allowed us to

reduce the Starlink satellite manufacturing cost per one Gbps of downlink capacity by approximately three times

from Starlink V1 Broadband satellites to V2 Mini satellites. We expect to achieve a total cost reduction of nine times

from Starlink V1 Broadband satellites to V3 satellites.

Starlink Broadband V2 and V3 Satellites

On Earth, users access the network through proprietary Starlink terminals that we design and manufacture. As of

March 31, 2026, we have reduced the cost of Starlink terminals—achieving an approximately 59% reduction in the

average manufacturing cost of a Starlink Kit since 2022—while improving performance and reliability, which we

believe collectively provides us a meaningful and durable competitive advantage over other terrestrial and satellite

broadband providers. Our portfolio of terminals, which we are able to manufacture and sell for a fraction of the cost

of terminals used by other satellite internet providers, includes three primary consumer configurations including: a

Standard terminal designed for fixed residential and small business use, featuring a wide field of view; a Mini

terminal roughly the size of a laptop, designed for mobility and travel use cases with a built-in Wi-Fi router and the

ability to operate on portable battery systems or 12V vehicle power; and the Performance terminal, designed for

demanding environments, with a maximum download speed over 450 Mbps and a higher power consumption of

190

110W or more under load. Each type of terminal is designed to be quick and seamless for a consumer to self-set up,

support in-motion connectivity up to speeds of 100 mph, and deliver global, oceanwide coverage for consumer

maritime use. We believe that this combination of low cost, portability (particularly in the case of our Starlink Mini

terminal), and ease of installation of our terminals will help scale our consumer broadband offering.

Starlink Standard and Mini User Terminals

We monetize Starlink primarily through subscription plans paired with hardware sales. Service tiers vary by speed,

priority access, geographic coverage, and mobility requirements, including Local and Global Priority options for

small to medium sized business, enterprise, and government Starlink customers. As the constellation scales and

capacity expands with next-generation satellites, we expect Starlink to continue growing as a global, recurring-

revenue connectivity platform and foundational layer of a space-enabled digital economy.

Enterprise Solutions

Enterprise Solutions offers the same fundamental advantages of Starlink Consumer Broadband—high throughput,

low-latency, and global coverage—into mission-critical, in-motion, and distributed connectivity environments for

enterprises. Starlink’s architecture is designed to deliver consistent performance across routes, oceans, and remote

industrial sites. Enterprise services are supported by dedicated hardware configurations and commercial structures

tailored to usage intensity, service-level requirements, and fleet-scale deployments.

191

Enterprise Solutions

Aviation Connectivity

Starlink Aviation provides broadband connectivity for commercial and private aircraft, enabling high-quality

internet service for passengers and crew from gate to gate, including during taxi and prior to take-off. The service is

differentiated by materially lower latency and higher throughput than legacy in-flight connectivity systems, enabling

streaming, video conferencing, and real-time applications at scale while in flight—even bandwidth-intensive

applications such as gaming, previously impractical from an airplane. Starlink’s global network is designed to

eliminate “dead zones” and supports performance on polar and high-latitude routes that can be challenging for

traditional providers. In recent years, we have assembled dedicated sales and engineering teams to market and

support fleet-wide conversions in the aviation sector. This has enabled partnerships with many of the world’s

leading airlines, including United Airlines, Southwest Airlines, Qatar Airways, Lufthansa Group, British Airways,

Alaska Airlines, and Hawaiian Airlines, many of which have implemented or committed to fleet-wide Starlink

installations for seamless in-flight connectivity.

Maritime Connectivity

Starlink Maritime provides broadband connectivity for vessels operating in coastal and deep-ocean environments,

supporting both operational requirements (navigation, telemetry, maintenance, logistics) and end-user connectivity

(crew welfare and passenger internet). The service is designed for consistent coverage regardless of proximity to

land, including routes that may experience service degradation under legacy satellite architectures. Starlink terminals

are engineered for marine operating conditions and are designed to be installed or swapped efficiently alongside

existing onboard communications systems, reducing downtime during retrofit. For many maritime operators,

Starlink functions as a wholesale or “syndicated” connectivity layer: vessel owners or cruise operators purchase and

allocate capacity across passengers, crew, and critical ship systems, including when reselling Wi-Fi access as an

onboard service. Pricing structures vary by vessel class, expected consumption, coverage requirements (coastal vs.

ocean), and priority level, and are generally implemented through recurring subscription arrangements with fleet-

based commercial terms. To support fleet-wide conversions in the maritime sector, we have partnered with premier

cruise operators, such as Carnival Corporation, Royal Caribbean Group, MSC Cruises, and Norwegian Cruise Line

Holdings, for full-fleet deployments that deliver reliable high-speed internet across thousands of vessels worldwide.

192

Land Mobility and IoT

Starlink supports in-motion connectivity for land mobility and industrial IoT applications where terrestrial networks

are intermittent or unavailable. These deployments include fleet vehicles, remote field operations, and ruggedized

use cases that require continuous broadband while moving, often across large geographies. The service is

particularly relevant for emergency responders, disaster recovery, and critical infrastructure continuity, where

resilient communications materially impact safety and response effectiveness. In industrial settings, Starlink can

serve as a connectivity backbone for connected equipment and telemetry-driven workflows, enabling real-time

monitoring and remote operations in agriculture, energy, and logistics environments. Commercial deployments are

typically structured around fleets or enterprise accounts, with hardware and service tiers aligned to mobility

requirements, usage intensity, and priority performance. We have partnered with land mobility operators, including

John Deere and the California Fire Department, as well as passenger rail operators such as Brightline (Florida), and

Italo Treno, to provide remote monitoring and management of their fleets.

Starlink Fixed Site

Starlink Fixed Site is designed to provide primary or backup connectivity for distributed business locations globally,

including sites that are difficult to serve economically with fiber or that require redundancy for uptime. Starlink’s

lack of dependence on wireline infrastructure—which is subject to damage or disruption from natural disasters,

conflict, and other events—makes it well-suited for businesses that rely on continuous broadband connectivity and

cannot afford a terrestrial offering going temporarily “offline.” Customers deploy Starlink to support point-of-sale

systems, corporate networking, video and security systems, and business continuity, including during disasters and

localized outages where terrestrial infrastructure may be impaired. The service is differentiated by rapid

installability, geographic flexibility, and reliable performance in remote and hard-to-reach locations, making it

suitable for retailers, industrial operators, and remote facilities (including offshore and field sites). Pricing models

include multiple tiers and configurations depending on speed, priority access, coverage footprint, and the number of

sites deployed, with typical enterprise arrangements structured as recurring subscriptions paired with hardware.

Government Solutions

We provide U.S. civil, state, and local government agencies as well as international civil government agencies high-

speed, resilient connectivity for public services, social impact, humanitarian efforts, and disaster response in even

the most remote and challenging environments. Examples include support for the FEMA in coordinating disaster

recovery after hurricanes and wildfires, the NOAA for at-sea testing and environmental monitoring, the Government

of the Philippines for linking remote islands, schools, and public institutions, the Government of Jamaica for

improving digital access in remote and maritime areas, and the Government of Ecuador for supporting education and

healthcare connectivity in isolated communities.

Separately, we operate Starshield, a secure satellite network designed specifically for national security applications.

Built on the technology, manufacturing, and launch infrastructure that underpin Starlink, Starshield is focused on

three core mission areas: Earth observation, global secure communications, and hosted payloads. Starshield satellites

are designed to integrate a wide range of sensors and instruments, allowing government customers to deploy

mission-specific capabilities in LEO without having to design, build, and launch standalone spacecraft for every

program.

Starshield builds on the end-to-end data encryption used in our commercial network by adding high-assurance

cryptographic capabilities tailored to military and other government requirements. By combining this security

posture with our high-cadence launch capability and evolving Starlink-derived infrastructure, we aim to offer a

scalable national security platform that can be updated, replenished, and expanded as mission needs change over

time.

Starlink Mobile

We are extending the reach of Starlink beyond fixed and mobility terminals through our mobile service, connecting

smartphones (with no modifications or incremental hardware) and other terrestrial devices directly to our satellites.

We aim to entirely eliminate mobile “dead zones.” By using satellites that effectively function as cell towers in

193

space, we enable data, over-the-top voice, video and messaging in remote and hard-to-reach locations where

terrestrial networks have historically been unavailable or unreliable. Starlink Mobile is already commercially

available for messaging in select markets and has been used to support emergency communications following

natural disasters, demonstrating its strength as resilient, infrastructure-independent connectivity.

V1 Mobile Satellites and V2 Mobile Satellites

Our mobile constellation builds on the same LEO architecture as our broadband network, with satellites specifically

designed to communicate directly with everyday LTE handsets and IoT devices without requiring specialized or

additional hardware. These satellites use exclusive licensed spectrum, allowing us to integrate into MNOs’ existing

networks while delivering coverage far beyond the reach of ground-based towers. Since launching the first mobile

satellites in early 2024, we have rapidly scaled the network to hundreds of in-orbit spacecraft and demonstrated key

technical milestones, including the first SMS tests within days of launch, live video calls, and public posts sent

directly from standard smartphones through a Starlink Mobile satellite. Our ability to design, manufacture and

launch these satellites on our own vehicles enables us to iterate quickly on payloads and software, expanding

capacity and performance over time.

Today, our Starlink Mobile service is delivered in partnership with leading mobile network operators around the

world. We are initially focused on messaging for consumer subscribers in areas with limited or no terrestrial

coverage, with a roadmap to support broader data, voice and IoT services. We partner with approximately 30 MNOs

across six continents, including T-Mobile in the United States, and other international operators including One NZ,

Optus, Telstra, Rogers, KDDI, Salt, Entel, Kyivstar, and VMO2. Through these partnerships, we enable consumers,

businesses and public-sector customers to use their existing phones in more places, support critical connectivity

during disasters and power outages, and open new applications for low-bandwidth mobile and IoT devices.

194

Map of Starlink Mobile Coverage

Satellite Life

We estimate that our satellites have useful lives of three to five years based on engineering studies, historical on-

orbit performance, propellant life, utilization patterns, design enhancements across generations, and planned

transitions to newer satellite technology. We, however, often deorbit satellites before the end of their useful lives,

primarily to reduce degradation risks that could impair our autonomous collision avoidance system and compromise

constellation safety. To date, our autonomous collision avoidance system has not experienced any failures resulting

in satellite loss, and satellite losses from other causes remain de minimis.

AI

Grok

Grok represents a core pillar of our mission to advance humanity’s understanding of the universe through the

development of truth-seeking artificial intelligence. Grok is designed and optimized for rigorous reasoning, real-time

information synthesis, and transparent outputs, with a product philosophy centered on intellectual honesty, first-

principles thinking, and engagement with complex topics.

Grok is designed as a truth-seeking AI model, built on our founder Elon Musk’s mission to enable humanity to

understand the universe. We believe that accomplishing this mission requires a truth-seeking approach to AI. We

define truth seeking as the active, relentless pursuit of what is objectively true about reality, and grounded in

evidence, logic, empirical data, and first principles thinking. Our goal is to understand and explain what the universe

appears to be doing, as accurately as current knowledge allows. In pursuit of this truth-seeking objective, Grok also

benefits from its integration with X, our real-time information, entertainment, and free speech platform. This direct,

real-time access to the information and human discourse on X enhances Grok’s truth-seeking capabilities by

grounding outputs in up-to-date knowledge and diverse viewpoints.

Since the initial release of Grok 1, we have iterated rapidly, releasing Grok 2, Grok 3, and, the current version, Grok

4, each delivering material improvements in pre-training, reasoning depth, multimodal capabilities, latency, and

scale. Building on this trajectory, we expect to continue scaling Grok through subsequent generations. Ongoing

195

training of next‑generation models is expected to scale toward multiple trillions of parameters, which could

represent a step change in reasoning in depth and overall intelligence. In this context, the number of parameters

refers to the scale of the model, where parameters are the internal numerical values, such as “weights,” that are

adjusted during training to enable the model to recognize patterns and relationships in data. A larger number of

parameters generally allows the model to capture more complex relationships, store greater amounts of knowledge,

and achieve higher levels of reasoning capability. Our accelerated development cadence positions Grok among the

fastest-advancing frontier models relative to peers, including OpenAI, Anthropic, and Google. Grok is differentiated

by its emphasis on real-time data integration, particularly through insights derived from the X platform (subject to

some limitations for certain content), enabling dynamic awareness of current events and user discourse, as well as by

explicit investment in reasoning transparency and explainability. Grok enhances the X ecosystem by improving

content understanding, personalization, and recommendation systems, thereby increasing user engagement and

platform intelligence. We are currently developing next-generation iterations, including Grok 5, which are expected

to further expand reasoning fidelity, multimodal integration, and domain-specific performance.

Terrestrial AI Compute

Our terrestrial AI compute forms the backbone of the Grok model family and is anchored by the COLOSSUS and

COLOSSUS II data centers that boast some of the world’s largest and most advanced AI training clusters.

COLOSSUS and COLOSSUS II collectively provide approximately 1.0 gigawatt of compute power, with the

additional power capacity available for data center operations. We brought the first cluster of COLOSSUS online in

122 days, repurposing the shell of an existing factory, and the first cluster of COLOSSUS II online even faster in 91

days. As an illustrative comparison, an industry benchmark to bring online a 100 megawatt greenfield data center is

approximately two years. We also demonstrated a significant improvement in cost efficiency, achieving data center

construction costs for COLOSSUS II that are considerably lower than industry benchmarks on a per megawatt basis.

COLOSSUS II is capable of operating entirely by our self-built behind-the-meter gigawatt-scale natural gas power

plant. Our data centers are integrated with the world’s largest Megapack deployment, providing additional layers of

reliability and operating performance. At all our existing data centers we have employed a brownfield retrofit

strategy leveraging existing industrial sites, advanced direct-to-chip cooling to support higher rack densities, and

high-speed networking. The clusters deploy leading-edge GPUs to maximize training throughput and model

performance. The next phase of expansion at COLOSSUS II is designed to train our next-generation Grok 5 AI

model. As we continue to expand our AI compute infrastructure, we will also continue to enhance our power

capabilities utilizing a combination of grid-power and behind-the-meter natural gas power plant buildouts. At

COLOSSUS, our grid power capabilities are designed to purchase power from the grid as available, and to rely on

our behind-the-meter, self-generated power and Megapack installations when grid power is curtailed.

196

COLOSSUS II Facility

X Platform

X is a real-time information, entertainment, and free speech platform that serves as a foundational distribution and

data engine for our AI ecosystem. With a global user base generating substantial volumes of content at all times

across a wide variety of topics, X provides a uniquely dynamic data for model training and real-time context

integration, subject to some limitations for certain content, which significantly differentiates Grok from the other

frontier lab offerings.

197

X is Our Real-time Information, Entertainment, and Free Speech Platform

X is our real-time information, entertainment, and free speech platform that serves as a global town square with

integrated AI capabilities powered by Grok. Designed to evolve toward an “everything app,” X enables users to post

content, share media, engage in conversations, host, view, and participate in live group discussions, follow real-time

events, use encrypted messaging, and leverage advanced features such as Grok-assisted post creation, content

discovery, and conversational AI directly within the interface via the prominent Grok icon.

198

Grok Holds Front and Center Real Estate on the X Platform

With native integration of Grok’s frontier models, including real-time access to X data for up-to-date insights,

trending analysis, and enhanced search, X delivers personalized feeds, smarter recommendations, and low-latency

AI assistance for our users worldwide. Our X Premium subscription options, including Basic, Premium and

Premium+ tiers, offer expanded features, ad-reduced experiences, and priority Grok interactions. In 2023, Grok’s

chat functionality was integrated into the X app allowing for the user to open the chat interface to type prompts and

get real time answers.

Public X data enhances Grok’s training and reasoning capabilities, while the platform continues to deliver

measurable performance outcomes for advertisers, with an increasing strategic focus on performance-based

marketing solutions.

In addition to X consumer products, X offers advertisers and developers a powerful suite of tools to reach highly

engaged audiences. Advertisers can target audiences through diverse ad formats—such as Promoted Ads, Vertical

Video Ads, Collection Ads, and premium options such as X Amplify and Takeovers—blending seamlessly with

organic content for authentic engagement. With advanced targeting based on public conversations, events,

keywords, interests, locations, and look-alike audiences, brands can connect with audiences while benefiting from

flexible, performance-based pricing (pay only for actions such as clicks or engagements) and often lower costs

compared to other platforms. We expect that our ongoing innovations—including Grok-powered integrations, new

contextual ad tests, and expanded aspect ratio support for easy reuse of ad creative—make X a competitive choice

for driving traffic, conversions, and brand awareness and visibility among X’s hundreds of millions of MAUs.

Developers have access to a continuous, high-volume, real-time stream of data around current events, trends, or

sentiment, which they can access through an official X Developer Platform and APIs.

In April 2026, we began a phased roll-out of our new advertising platform, that we rebuilt from the ground up. The

new Ads Manager is built to help advertisers launch better campaigns, faster, with stronger ROI. Powered by AI, the

new systems enable more precise, relevant and dynamic ad delivery. Ads are seamlessly integrated into a User’s X

feed.

199

By combining high-volume user interactions with frontier AI, AI compute infrastructure, and vertical integration, X

accelerates progress toward ubiquitous connectivity, real-time global awareness, and the foundational social layer

for multiplanetary human endeavors.

X Ads Manager. X provides a comprehensive suite of advertising products, including promoted posts, video ads,

carousels, and sponsored content, which enable businesses to reach targeted audiences in real time across the

platform. Powered by real-time conversation data, interest-based targeting, and behavioral signals, these solutions

support objective-based campaigns focused on website traffic, video views, app installs, lead generation, and brand

awareness. X’s Ad Manager provides a centralized platform that allows advertisers to manage creation,

optimization, and real-time monitoring of ad campaigns with detailed audience insights, bidding controls,

performance analytics, and A/B testing capabilities. Integration with Grok AI further streamlines creative

development, making X’s scalable ad solutions effective for businesses of all sizes seeking efficient engagement in a

dynamic public conversation environment.

Grok Consumer Products

Our consumer products are powered by Grok, including Grok language and coding models, Grok image and video

generation models (more commonly known as Grok Imagine), and Grokipedia. These applications leverage the

underlying Grok model family to deliver advanced multimodal interaction, real-time information awareness, and

transparent reasoning outputs. We currently offer three different tiers of subscription for Grok—basic, SuperGrok,

SuperGrok Heavy, and SuperGrok Lite, each priced on a monthly or annual basis. Higher pricing tiers unlock

expanded access to advanced models, increased usage limits, priority compute, and a suite of premium features

tailored to power users and enterprise-grade applications.

Grok Chat. Grok Chat represents the primary conversational interface of Grok, enabling users to submit text or

voice queries for explanations, problem-solving, research, coding, brainstorming, and in-depth discussions with real-

time integration of web search, X data, code execution, and multimodal analysis of images or documents. Available

via grok.com, dedicated mobile apps, X platform integration, and the xAI API, it provides truth-seeking, helpful,

and minimally censored responses optimized for factual precision and complex reasoning.

Grok Chat

200

Grok Imagine. Grok Imagine is Grok’s generative visual and multimedia creation suite, powered by proprietary

models for producing high-quality images, short videos (up to 15 seconds at 720p in current iterations), and

synchronized audio from text prompts, reference images, or existing visuals. It supports text-to-image/video editing,

image-to-video editing, and video-to-video editing, style transfer, and cinematic motion with strong prompt

adherence and photorealistic output, accessible through the Grok platform, Imagine tab, and dedicated API.

Grok Imagine

Grok Voice. Grok Voice delivers natural, real-time conversational AI through voice interactions, allowing users to

seamlessly speak and listen to Grok for faster access to information and task execution.

Grok Enterprise Products

Grok Teams. Grok Teams empowers small-to-medium-sized organizations to integrate Grok’s advanced AI

capabilities directly into collaborative workflows. Teams gain access to dedicated workspaces with secure sharing,

enhanced privacy protections, and administrative controls for inviting users and managing access. Grok Teams

accelerates analysis, innovation, and creation while ensuring data remains private and is never used for training.

Grok API. The Grok API provides programmatic access to Grok’s frontier models, including advanced reasoning,

vision, tool-use, image generation, voice AI, and real-time search capabilities, tailored for enterprise-scale

integration. It offers features like agentic workflows, and enterprise-grade options such as custom allocations, secure

authentication, and dedicated support. Designed for developers and organizations building production applications,

the API enables seamless embedding of Grok’s powerful AI into custom solutions, driving innovation across

industries with speed, precision, and reliability. For example, the enterprise version of the Grok Voice Agent API

allows developers and businesses to build multilingual voice agents capable of speech recognition, tool calling, real-

time data querying, and low-latency responses. It supports production-grade voice applications that enhance

customer service, internal operations, and interactive experiences with high performance in audio reasoning

benchmarks.

201

Infrastructure and Facilities

SpaceX maintains a highly vertically integrated, geographically diverse manufacturing ecosystem that designs,

produces, and qualifies a significant share of components in-house, from raw materials and rocket engines to

complete launch vehicles, crewed spacecraft, satellites, and user terminals, enabling unprecedented iteration speed,

quality control, and cost efficiency essential for successful production of reusable systems and high-cadence

operations. Our manufacturing facilities are complemented by our physical infrastructure, which supports launch

and orbital operations for human spaceflight, satellite deployment, and cargo missions, as well as large-scale

artificial intelligence training and inference. We continue to invest in expansions and improvements across our sites

to accommodate anticipated growth in launch cadence, Starlink Subscribers, and AI compute requirements.

SpaceX Facilities

While none of our properties are individually material to our operations because of the long-term timetables for

renewal and the opportunities for alternative sites, we maintain an effective network of vertically integrated facilities

across the United States, including:

•Starbase, Texas: Development, manufacturing, testing, and launch of Starship currently takes place at

Starbase, home to SpaceX headquarters and one of the world’s first commercial spaceports designed for orbital

missions. The site is located at the newly created city of Starbase in Cameron County, Texas, along the Gulf of

America. Its infrastructure includes Starfactory, a manufacturing facility designed to mass produce Starship and

Super Heavy at scale; a large office structure co-locating engineering and production personnel; and large,

vertical integration buildings including the upcoming Gigabay, which will be able to support Starship and Super

Heavy vehicles up to 85 meters (279 feet) tall and will provide 24 work cells for integration and refurbishment

work, along with cranes capable of lifting up to 400 tons. Starbase also has an orbital launch pad for flight of the

world’s most powerful rocket, complete with one of the tallest launch towers in the world, specially designed to

integrate, test, launch, and catch Starship and Super Heavy vehicles, with an additional pad underway to support

Starship V3. The Starbase team also operates a site for full and subscale vehicle structural testing, static fires,

and component level testing.

Starbase is also home to several hundred SpaceX employees and their families, many of whom have relocated

from across the country to the community to support the development and operation of Starship. SpaceX, in

202

partnership with the newly formed city, is developing local infrastructure and municipal services, including

utilities, governance, schools, and environmental conservation initiatives, to support a world-class, concentrated

engineering and manufacturing community focused on the rapid advancement of Starship and SpaceX’s long-

term mission. This close integration of residential life, engineering, and manufacturing around a single program

enables a mission-focused environment designed to accelerate development, testing, and launch operations.

SpaceX Headquarters at Starbase, Texas

203

•Hawthorne, California: Our original flagship facility in Hawthorne, California manufactures Falcon 9 and

Falcon Heavy first and second stages, Dragon Crew and Dragon Cargo spacecraft, Merlin engines, Starship’s

Raptor engines, Starlink User Terminals, as well as other various Starship components. The site supports high-

reliability production for hundreds of successful missions, including NASA-certified crew rotations. We also

maintain a corporate presence in Hawthorne.

Hawthorne, California

204

•McGregor, Texas: The McGregor rocket engine complex is the most active rocket development and testing

facility in the world. It serves as the primary site for qualification, acceptance, and post-flight testing of Merlin

and Raptor engines. It features 15 specialized test stands, including dedicated vertical stands for Raptor engines

and multiple stands for Falcon 9’s Merlin engines, as well as component-level testing facilities for Starship

hardware, including composite overwrapped pressure vessels, tanks, and experimental systems.

McGregor, Texas

205

•Redmond, Washington: TheRedmond Starlink satellite manufacturing facility has produced an average of

approximately 70 satellites per week (approximately 3,640 per year at full rate) from December 2025 to April

2026, covering bus structures, phased-array antennas, propulsion, solar arrays, and inter-satellite lasers,

enabling rapid Starlink constellation expansion.

Redmond, Washington

206

•Bastrop, Texas: We build the majority of Starlink products at our manufacturing facility in Bastrop, Texas,

which opened in 2023, producing tens of thousands of Starlink Kits per day and all of the current generation

Starlink Standard and Performance Kits.

In 2026, we expect to more than double the size of the Bastrop facility, expanding our design and

manufacturing capabilities to support new Starlink products, plus deepening our vertical integration by adding

the production of Starlink gateway antennas, solar cells and AI compute satellites.

Bastrop, Texas

•Kennedy Space Center and Cape Canaveral, Florida: SpaceX operations in Florida span across NASA’s

Kennedy Space Center and Cape Canaveral Space Force Station, which includes two active launch sites—

Launch Complex 39A (LC-39A) and Space Launch Complex 40 (SLC-40)—Falcon booster and Dragon

spacecraft refurbishing facilities, launch operations, and payload processing buildings. Both launch sites support

critical missions to geostationary orbit and the International Space Station while also providing launch

opportunities to a wide range of low, mid, and polar orbit inclinations for science and national security

missions. SpaceX also utilizes Landing Zones 40 and 2 at the Cape, which support Return to Launch Site

landings for Falcon boosters ahead of recovery and refurbishment for future missions.

Once recovered, flight hardware is refurbished at one of two state-of-the-art SpaceX facilities, HangarX and X2,

on Kennedy Space Center. These facilities also house our Falcon Launch and Landing Control Center, where

our Dragon spacecraft are refurbished and prepared for their next missions after they are recovered off the coast

of southern California, where we produce Starship heatshield tiles in the Bakery, and where we process

customer payloads before launch in our Payload Processing Facility.

For future launches, SpaceX is expanding its operations in Florida to bring Starship to the Cape. In addition to

the under-construction Starship launch pad at LC-39A expected to be completed by the end of 2026, SpaceX is

constructing Space Launch Complex 37 (SLC-37) on Cape Canaveral Space Force Station as another Starship

launch site. SLC-37 will host two orbital launch pads, including up to two towers for Starship launch, catch, and

testing operations, culminating in a total of four operational launch pads for Starship by the end of 2027.

SpaceX is also building a new integration facility called Gigabay, next to its HangarX location at Kennedy

Space Center by late 2026.

207

In connection with preparing leased real property for our launch operations, we make significant capital

improvements and install extensive real and personal property at these government-owned sites. The launch

facilities we build are a unique capital improvement compared to standard commercial use sites because the

federal government specifically designates these launch sites for aerospace activities, such as rocket launches.

Given the specific use requirements of these government-owned sites, we have historically entered into

handover agreements with the relevant government entities upon expiration or termination of the leases,

pursuant to which the improvements are transferred to the government rather than removed. This fact pattern

has historically been the case with previous leases such as at Cape Canaveral Space Force Station.

NASA’s Kennedy Space Center, Florida

208

Cape Canaveral Space Force Station, Florida

•Vandenberg Space Force Base, Space Launch Complex 4: Space Launch Complex 4 East at Vandenberg

Space Force Base is our West Coast launch site and serves as our primary facility for polar and high-inclination

orbit missions critical to Starlink constellation deployment, national security payloads, Earth observation

satellites, and select lunar trajectories. The facility includes a modernized orbital launch pad optimized for

Falcon 9 launches, featuring a fixed launch mount, integration tower, propellant loading infrastructure, flame

trench, and support systems enabling frequent operations. Adjacent Space Launch Complex 4 West functions as

a dedicated Falcon 9 booster landing zone, supporting downrange recoveries to maximize reusability. Please

refer to “—Kennedy Space Center and Cape Canaveral, Florida” for additional information regarding our lease

arrangements with government entities.

209

Vandenberg Space Force Base, California

210

•Memphis, Tennessee and Southaven, Mississippi: We operate a cluster of high-density data centers in the

Greater Memphis Area extending into northern Mississippi along the state border, to power training and

inference for frontier AI models, including the Grok family. The flagship COLOSSUS supercomputer campus

is located on Paul R. Lowry Road in Memphis, Tennessee; the COLOSSUS II facilities are located on Tulane

Road in Memphis, Tennessee and on Stateline Road in Southaven, Mississippi.

Memphis, Tennessee

211

•Palo Alto, California: The corporate headquarters for our AI operations following the acquisition of xAI in

February 2026 is located in Palo Alto, California. This location, under long-term lease, houses our advanced AI

research, development, and engineering teams and is strategically situated in Silicon Valley to attract and retain

top AI research talent. The engineers responsible for the design, training, and continued evolution of Grok, our

proprietary frontier AI model, are based at this facility.

Palo Alto, California

In addition to our infrastructure and facilities across the United States, we also operate a fleet of recovery vessels,

autonomous spaceport drone ships (“ASDS”), and a network of Starlink ground stations.

•Our recovery fleet:Our fleet of ASDS forms the maritime backbone of SpaceX’s reusable rocket architecture,

enabling high-probability downrange booster landings for Falcon 9 and Falcon Heavy missions while

maximizing vehicle recovery and rapid refurbishment. The core ASDS fleet consists of three operational

vessels: “Of Course I Still Love You,” the pioneering East Coast-to-Pacific vessel homeported at the Port of

Long Beach, California, and dedicated to supporting primarily polar and high-inclination launches from

Vandenberg Space Force Base with its large landing deck and thruster-based dynamic positioning; “Just Read

the Instructions,” stationed at Port Canaveral, Florida, serving East Coast operations from Cape Canaveral and

Kennedy Space Center; and “A Shortfall of Gravitas,” the newest and most advanced addition since 2021, also

based at Port Canaveral with enhanced autonomy, station-keeping precision, and upgraded deck infrastructure

to handle frequent, high-cadence missions. These autonomous ships have collectively facilitated hundreds of

successful booster touchdowns, dramatically reducing expendable flight profiles and enabling the reuse of

boosters 34 times as of March 31, 2026. Complementing the drone ships are dedicated support vessels for

fairing half recovery, such as “Bob” and “Doug,” named after astronauts Bob Behnken and Doug Hurley, and

Dragon retrieval vessel “Shannon,” named in honor of astronaut Shannon Walker. These support vessels ensure

comprehensive ocean-based recovery operations across Atlantic and Pacific theaters and underpin our

constellation deployments, national security launches, and crewed missions while advancing toward full

reusability for Starship in future offshore scenarios.

212

Autonomous Drone Ship “A Shortfall of Gravitas”

•Starlink ground stations: A Starlink ground station, also referred to as a gateway, is a terrestrial relay station

that communicates with our satellite constellation. These stations transmit data between satellites and terrestrial

internet networks. We operate ground stations around the world, with over 400 sites globally.

Customer Case Studies 

The following examples illustrate ways in which customers across a range of industries have used and benefited

from our solutions within our Space, Connectivity, and AI segments. These examples are intended to highlight

representative applications of our offerings and the types of operational, performance and efficiency benefits that

customers may realize.

In addition, we include examples of our deployment of Starlink services in response to natural disasters, which

demonstrate our ability to rapidly establish communications infrastructure to support emergency response and

recovery efforts in challenging environments.

219

Competition

Our principal sources of competition vary based on the segment and market in which our business operates.

In Space, we compete with launch service providers that transport small, medium, and heavy payloads and

astronauts to Earth’s orbit and beyond. Participants in this market include established aerospace and defense

companies, emerging commercial launch providers, and national space agencies. Key established aerospace and

defense competitors providing launch services include, among others, United Launch Alliance, a joint venture

between Boeing and Lockheed Martin, Arianespace, a French-based aerospace company operating a family of

European-developed rockets, and Northrop Grumman, manufacturer of the Cygnus cargo spacecraft. Emerging

commercial launch providers include Blue Origin, which has developed launch vehicles intended to compete with

our Falcon 9 rocket, and Rocket Lab, which operates in the small-lift launch market but is expanding into medium-

lift payloads, as well as other domestic competitors such as Firefly Aerospace and Relativity Space. While we

typically do not compete directly for the same missions, national space agencies also provide launch services in their

respective markets.

However, the launch services market is characterized by significant barriers to entry, including substantial capital

requirements, advanced technological expertise, regulatory licenses and approvals, and established relationships

with government and commercial customers. Competition in this market is based on factors that include launch

reliability and cadence, payload capacity, mission flexibility, manufacturing capabilities and price. For this reason,

while the established aerospace and defense competitors and emerging commercial launch providers may provide

launch services at varying degrees of scale, we believe that SpaceX holds a meaningful advantage in terms of the

breadth of our launch solutions and services and the cadence at which we are able to launch, and thus a significant

competitive advantage relative to these players.

In Connectivity, we compete with operators of terrestrial and satellite communications infrastructure and providers

of satellite-to-mobile connectivity solutions, including terrestrial fixed network providers, terrestrial mobile network

companies, and other satellite service providers, as described below:

•Consumer and Enterprise Broadband. Our Starlink Consumer and Enterprise broadband offerings compete with

terrestrial fixed network providers, terrestrial mobile network companies, and other satellite service providers.

Terrestrial fixed network providers include operators of cable and fiber networks such as Verizon, Comcast,

AT&T, T-Mobile, Lumen, Charter Communications, Google Fiber, Astound, BT, Deutsche Telekom, and

Liberty Global. Terrestrial mobile network companies also operate land-based infrastructure, including wireless

antennas affixed to mobile towers used to provide fixed wireless services, and include AT&T, Telefónica, T-

Mobile, Verizon, and Vodafone Group. These network providers typically serve customers in one or more

countries (for example, Verizon in the United States, or Telefónica in Spain and Brazil, among others), but are

not global players insofar as they do not sell to a global customer base, nor does their network infrastructure

exist globally. Satellite service providers include, among others, GEO satellite network operators such as

EchoStar, SES, Telesat Corporation (“Telesat”) GEO, and Viasat, as well as current and planned LEO and

MEO constellations including Amazon LEO, Blue Origin’s TeraWave, Eutelsat OneWeb, Iridium NEXT and

Telesat Lightspeed. Some of these service providers are also launch customers of SpaceX as they contract with

us to launch their satellite constellations into orbit.

•Government Solutions. Our Starlink broadband offering for government use cases competes primarily with the

same terrestrial network providers and satellite service providers with which our Starlink Consumer and

Enterprise broadband offerings compete, as well as defense prime contractors. In certain cases, these providers

also have dedicated subsidiaries or business units focused on serving government customers, such as Telesat

Government Solutions.

•Starlink Mobile. Our Starlink Mobile offering competes with other satellite-to-mobile satellite operators

including, among others, AST SpaceMobile, Lynk, Globalstar and Skylo.

The satellite connectivity market involves significant barriers to entry, including substantial capital requirements,

advanced technological capabilities, access to spectrum and orbital resources, regulatory licenses and approvals, and

the development of relationships with government, enterprise and commercial customers. Competition in this market

220

is based on factors that include network coverage, capacity, latency and reliability, spectrum access, density of urban

environments, satellite deployment capability and efficiency, price and user acquisition, retention, and experience.

In AI, we compete with developers of foundational AI models and providers of AI products and services, as well as

general purpose and vertical search engines, information services, online advertising platforms and social networks.

Participants in this market include large technology companies, emerging AI model developers and providers of AI-

enabled products and services. Key competitors in these markets include, among others, AI model developers and

platform providers such as OpenAI, Anthropic, Google, Meta, Microsoft, and various open source model providers,

as well as social networks such as Threads (owned by Meta), Reddit, and TikTok. As we continue to build out our

AI compute infrastructure, we intend to sell our excess capacity by offering it to a limited number of third parties

and intend to continue to explore monetizing excess capacity, potentially positioning us to emerge as a competitor to

AI cloud providers such as Coreweave and Nebius as well as hyperscalers.

Our AI businesses likewise compete in markets characterized by significant barriers to entry, including substantial

computational and infrastructure requirements, access to large datasets and the ability to attract and retain highly

skilled technical talent. Competition in these markets is based on factors including pricing and cost efficiency, the

performance and technical features of AI platforms, customer experience across our products and services, the

ability to attract new and retain existing subscribers, users and advertisers and the ability to deploy compute and

innovative technologies at scale.

Intellectual Property

The intellectual property that is material to our business includes our proprietary knowledge and software, as well as

our brands and our selectively patented inventions and technologies. Our proprietary knowledge includes expertise

in design, testing, manufacturing, software, in-orbit operations, real-time platforms, and artificial intelligence

development. The protection of our technology and intellectual property is an important aspect of our business. We

rely upon a combination of patents, trademarks, trade secrets, copyrights, confidentiality procedures, contractual

commitments and other legal rights to establish and protect our intellectual property. We have registered, and

applied for the registration of, U.S. and international trademarks, service marks, domain names, and copyrights. We

have also filed patent applications and acquired patents in the United States and foreign countries covering certain

aspects of our technology, and in some cases, we have acquired patent assets of others to supplement our portfolio.

We have licensed in the past, and expect that we may license in the future, certain of our rights to other parties or

from other parties. We generally enter into confidentiality agreements and invention or work product assignment

agreements with our employees, contractors, and consultants to control access to, and clarify ownership of, our

proprietary information and other intellectual property. For additional information, please refer to “Risk Factors—

Risks Related to Our Business—We may face substantial potential liability and operational disruptions if we violate

the intellectual property rights or other rights of third parties, and if we fail to adequately protect, maintain, defend

or enforce our intellectual property and other similar rights, we could lose an important competitive advantage, in

each case which could have a material adverse effect on our business, financial condition, results of operations,

customer trust and future prospects.”

Human Capital

As of March 31, 2026, we employed over 22,000 full-time employees worldwide, none of whom are subject to any

collective bargaining agreement. We believe our strong culture of collaboration and innovation distinguishes us and

serves as an important driver of our business performance.

Regulatory Environment

We are required to comply with a variety of governmental regulations, which could have a significant impact on our

business, including our capital expenditures, earnings and competitive position. In particular, our ability to (i)

conduct launches and reentries, (ii) operate and expand our satellite systems and related ground infrastructure and

(iii) perform certain U.S. government programs depends on maintaining key governmental authorizations and

complying with evolving safety, spectrum, national security, environmental, contractual, and trade-control

requirements. Our ability to provide our AI products and X platform depends on complying with evolving AI, data

privacy, online services, cybersecurity and environmental requirements. We incur and will continue to incur

221

substantial costs to monitor and take actions to comply with governmental and other regulations that are or will be

applicable to our businesses, including, among others, restrictions and regulations of the U.S. Department of

Transportation, the FAA, the FCC and other government agencies in the United States and the other countries in

which we operate, economic sanctions and trade embargo laws, export controls, import controls and customs. For

additional information, please refer to “Risk Factors—Risks Related to Our Business—Our ability to continue and

expand launch and satellite operations depends upon our ability to obtain new and leverage existing U.S. export

control and sanctions authorizations, and any significant changes to the geopolitical landscape or U.S. government

regulatory approach to licensing could materially and adversely impact our international business operations by

compromising existing licenses or limiting our ability to engage in commercial dealings in or involving

geopolitically sensitive countries.” We will also be subject to additional laws and regulations as a result of being a

public company, which will require us to devote significant management resources and incur additional legal,

accounting and other expenses.

Space

Our Space segment is subject to extensive regulation in the United States and internationally, including (i)

regulations administered by the FAA relating to commercial space launches and reentries, (ii) regulations

administered by the FCC relating to radio communications used in launch activities and spacecraft operations, and

related domestic and international coordination processes, including through the International Telecommunication

Union, (iii) U.S. export and import regulatory regimes, and (iv) additional regulations that relate to being a U.S.

government contractor.

Commercial space launch and reentry activities require licenses and permits from the FAA. FAA licenses are

generally granted on a launch-by-launch basis and may incorporate safety, environmental and operational

conditions. Where applicable, reentry operations require separate authorization. We are generally required to obtain

licenses or license modifications from the FAA in connection with changes to vehicles, launch sites, flight profiles,

operational procedures, payloads, or other mission parameters, and our launch and range operations may also be

subject to environmental reviews, consultations, and permits. We depend on timely approvals of licenses or license

modifications from the FAA and the timing and outcome of the FAA approval process may affect our ability to

conduct launches and reentries or require operational restrictions or mitigation measures. For additional information,

please refer to “Risk Factors—Risks Related to Our Business—Any delays or difficulties in obtaining, maintaining

or renewing required regulatory approvals and licenses required for our space-related activities, including FAA

launch and reentry licenses, would materially delay or disrupt our operations, harm our business, or limit our ability

to execute our business strategy.”

Radio communications for launch activities and spacecraft operations require licenses from the FCC and are subject

to technical and operational conditions, coordination requirements, and interference-mitigation frameworks. We rely

on obtaining licenses from the FCC to conduct our launch and spacecraft operations, and many of our FCC licenses

include conditions regarding milestone schedules, reporting and surety‑bond requirements, among other conditions.

In addition, our spacecraft and satellite operations are subject to evolving regulatory expectations relating to space

situational awareness and orbital debris mitigation, including requirements regarding collision avoidance and post-

mission disposal. International spacecraft frequency use is coordinated via International Telecommunication Union

filings made through the FCC and similar international regulatory bodies, and through country‑by‑country market

access approvals for non‑U.S. service. For additional information, please refer to “Risk Factors—Risks Related to

Our Business—Any delays or difficulties in obtaining, maintaining or renewing required regulatory approvals and

licenses required for our space-related activities, including FAA launch and reentry licenses, would materially delay

or disrupt our operations, harm our business, or limit our ability to execute our business strategy.” 

Additionally, as a contractor and subcontractor to certain agencies of the U.S. government, we are subject to the

Federal Acquisition Regulation, and other applicable laws, security requirements, and regulations, including

supplemental agency regulations, which comprehensively regulate the formation, administration, and performance

under government contracts. Certain contracts with the U.S. government may require us to be issued facility security

clearances under the National Industrial Security Program Operating Manual Rule, as a result of which we are

required to maintain with the Department of War mitigation measures with respect to foreign ownership, control and

influence. Additionally, certain transactions in which we may be involved from time to time may be subject to the

222

jurisdiction of the Committee on Foreign Investment in the United States (“CFIUS”), which has authority to conduct

national security reviews of certain foreign investments. CFIUS may impose mitigation conditions to grant clearance

of a particular transaction, may unilaterally initiate national security review of certain transactions, and may

recommend that the President of the United States order parties to divest their shareholdings in certain situations,

among other actions.

Connectivity

Our Connectivity services, including our global satellite-to-mobile connectivity services under Starlink Mobile,

depend on authorizations from the FCC in the United States and telecommunications regulators in other countries.

Without these licenses and approvals, we generally cannot offer connectivity services in a given market. In the

United States, these authorizations include FCC approvals for our satellite system and related earth stations and use

of radio frequency spectrum, and they may be subject to technical, operational, and reporting conditions and

ongoing compliance obligations (including interference mitigation, coordination requirements and orbital debris

mitigation requirements). All communications services that rely on radio frequency communications require use of

radio frequency spectrum, the assignment and distribution of which is subject to FCC oversight. Our access to

spectrum and orbital resources is also subject to international coordination processes, including through International

Telecommunication Union filing and coordination processes, and disputes or delays in these processes could

adversely affect our operations. If demand continues to increase or if new spectrum is required for a future

generation of technology, we may need to obtain additional spectrum usage rights or related authorizations through

FCC proceedings (including modification applications), coordination processes, auctions or secondary market

transactions, or partnerships with third parties, each of which may be subject to review, approval, and conditions.

We hold FCC authorizations and licenses that allow us to provide a wide range of satellite-based connectivity

services, including through the operation of our satellite system and related earth stations. FCC spectrum licenses

and authorizations typically have terms of 10-15 years, at which time they are subject to renewal. Similarly, our

subsidiaries operating outside the United States are subject to the jurisdiction of regulatory authorities in the

territories in which the subsidiaries operate, including any requirements to obtain spectrum licenses or other market

access authorization. Our licensing, compliance and advocacy initiatives in foreign countries support our ability to

offer enterprise and consumer connectivity services in various international markets. Although we generally seek to

renew and maintain these authorizations, challenges could be raised in the future, and there can be no assurance that

our applications to renew, modify, or expand our authorizations will be granted on a timely basis, or at all, or

without additional conditions. If a spectrum license was revoked or not renewed, we would not be permitted to

provide services on the spectrum covered by that license or could be required to modify or curtail operations.

Within the United States, the Communications Act generally preempts regulation by state and local governments of

the entry of, or the rates charged by, wireless carriers. It does not prohibit states from regulating the other “terms and

conditions” of wireless service. For example, some states impose reporting and consumer protection requirements.

Several states also have laws or regulations that address safety issues (for example, use of wireless handsets while

driving), universal service funding, and taxation matters. Some states are also considering new network reliability or

service quality requirements that may affect how and where we provide services if not preempted by federal law.

AI

Certain enacted and proposed laws and regulations related to AI may impose requirements with respect to our

development, deployment, and use of AI systems and models, including obligations relating to security, integrity,

transparency, labeling, detection, and provenance of AI data, models and AI-generated content, as well as

restrictions on the export or import of AI-related systems and components. AI regulation is evolving rapidly across

jurisdictions, with regulators applying, or considering applying, existing laws or adopting new, non-harmonized

frameworks with respect thereto, including emerging AI laws. Development, deployment, and use of AI can also be

subject to existing, technology-agnostic regulatory frameworks, including, for example, those addressing consumer

protection, data privacy, cybersecurity, intellectual property, content moderation, non-discrimination, and

employment. Data centers necessary for AI-related systems may also be subject to changing regulatory frameworks

under federal, state, local, and foreign environmental, health, and safety laws. The scope and enforcement of these

regimes remain uncertain, and their potential impact on our multiple and overlapping business lines is difficult to

223

predict. Divergent or conflicting regulatory approaches across jurisdictions, as well as evolving enforcement

priorities, may also create compliance uncertainty and require market-specific limitations or modifications to AI-

related functionality, increasing operational complexity.

In addition, third parties may allege intellectual property violations, or misappropriation relating to the training data

used in, or the outputs generated by, AI systems and models. The uncertain and evolving legal status of AI-

generated content may create legal and operational risk, including with respect to the ownership of, and ability to

obtain intellectual property protection for, such outputs, as well as our ability to offer services in certain markets.

Open-source and other license terms applicable to AI systems and models may limit the distribution of AI-related

functionality or constrain product design.

Separately, AI systems and models may present legal operational and reputational risks. Legal and reputational risk

may arise in the context of datasets used in the development or operation of AI systems and models as well as the

use of AI-enabled products or services to generate output that is perceived as objectionable or inappropriate.

Emerging legislation, such as the European Union’s Artificial Intelligence Act, California’s Transparency in

Frontier Artificial Intelligence Act (SB 53) and New York’s Responsible AI Safety and Education Act (RAISE Act),

may impose requirements relating to, among other things, safety, governance, transparency, and incident reporting

on developers of large or frontier AI models. Misuse of our AI systems, models, products, or services by customers

or partners may similarly create safety, compliance, or brand risks. These risks have in the past and may in the future

result in regulatory scrutiny, legal liability, or reputational harm and adversely affect our business, results of

operations, and financial condition. Addressing these risks may require substantial investment in testing,

moderation, guardrails, enforcement, and other mitigation measures. For additional information, please refer to

“Risk Factors—Risks Related to Our Business—If the recommendations, forecasts, content, analyses or other output

that our AI technologies, including Grok, assist in producing are or are alleged to be deficient, inaccurate, harmful,

illegal, or used for an improper purpose, we could continue to be subjected to claims and investigations, and we

could be subjected to legal liability and brand, reputational, or competitive harm.”

Privacy, Cybersecurity, Data Protection, Online Safety, and Digital Platform Regulation

We are subject to complex and evolving global legal and regulatory frameworks relating to privacy, cybersecurity,

AI, data protection, lawful access, content moderation, and digital platform regulation, as well as contractual and

other commitments we make in the course of doing business and our internal and external policies, procedures and

controls. These laws and regulations vary across jurisdictions and sectors, are not harmonized, and may conflict or

impose overlapping or inconsistent obligations, and continue to evolve and emerge. In particular, the California

Consumer Privacy Act (as amended), the European Union’s General Data Protection Regulation (and its equivalent

in the United Kingdom) and other data privacy laws and regulations impose stringent and burdensome requirements

in connection with the processing of personal information and include significant penalties for non-compliance.

Additionally, as a government contractor, we are also subject to the Department of War’s Cybersecurity Maturity

Model Certification requirements, which requires companies that do business with the Department of War to,

depending on the level of security required, meet or exceed certain specified cybersecurity standards to be eligible

for new contract awards. The interpretation and application of these and other existing laws not originally enacted to

address privacy, cybersecurity, AI, data protection, lawful access, content moderation, or digital platforms are

uncertain and continue to develop as they are applied to new technologies and data-driven products and services.

These frameworks impose obligations regarding, among other things, the collection, use, storage, protection,

disclosure, transfer, and other processing of data, including personal information, and may restrict or condition

cross-border data transfers, require data localization, or impose content moderation or other platform-related

requirements, and may be interpreted or enforced in ways that are inconsistent, unclear, or subject to significant

regulatory discretion. The risks are particularly acute for us because we operate globally across multiple industries

and develop cutting-edge technologies that present novel regulatory and security issues. The data we collect and

otherwise process is integral to our business, technology, and services, and regulatory restrictions or limitations on

our ability to secure and process such data could materially affect our operations and business model.

In addition, our products and services, including those enabled by AI, may also be subject to online safety and

youth-protection laws and regulations. Such laws and regulations may impose obligations relating to content risk

mitigation, age assurance, platform governance, and, in certain jurisdictions, content reporting and removal

224

requirements. For example, the UK’s Online Safety Act 2023 and Australia’s Online Safety Amendment (Social

Media Minimum Age) Act 2024 impose risk mitigation and age-related requirements on certain online platforms. As

a result of these requirements or to otherwise seek to maintain the safety of our platforms, we maintain content

policies and enforcement mechanisms across our platforms and related products and services. These include a

combination of automated detection tools, classifiers and filters, algorithmic signals, and human review processes.

We also employ measures to help detect and challenge suspicious accounts during sign-up and ongoing use, provide

user reporting channels, and apply enforcement actions. Additional safeguards to help mitigate safety concerns

include age-related controls, content restrictions, and specialized modes; and labeling or watermarks on certain

outputs and other market-specific restrictions on certain content categories where required by local laws.

This evolving landscape will continue to affect our ability to maintain, develop, or launch products and services,

including those that rely on the processing of personal information or other sensitive data, including targeted

advertising and other data-driven offerings, and may require market-specific changes to our products, services, or

business practices, increasing operational complexity and cost. In addition, emerging laws and regulations seeking to

restrict cross-border transfer of or access to certain data in light of perceived national security considerations may

increase compliance costs and restrict our operational flexibility, investment activities, or ability to achieve our

strategic objectives. As our business evolves, and if we expand into additional industries or jurisdictions, our

compliance requirements and associated costs may increase and we may be subject to heightened regulatory

scrutiny.

We also face cybersecurity risks, including the potential unlawful, accidental, or unauthorized access to, or use,

disclosure, alteration, loss, or disruption of, our technology, products, systems, and data, or those of our service

providers and partners, which could result in a loss of confidentiality, integrity, or availability. We operate in

industries that have been, and will continue to be, targeted by sophisticated and persistent internal and external threat

actors, including those controlled by or affiliated with nation states. For additional information, please refer to “Risk

Factors—Risks Related to Our Business—Any significant disruption in, or unauthorized access to, our computer and

data systems or those of third parties that we utilize in our operations could result in a loss or degradation of service,

loss of trust in us and harm to our business.” Many jurisdictions impose mandatory breach notification and reporting

obligations, and compliance with such requirements can be costly, time-sensitive, and operationally burdensome,

and we may bear such costs in the event of a material incident. As we continue to use and integrate advanced

technologies, including AI systems and models, into our operations, products, and services, our exposure to

cybersecurity incidents may increase, particularly as threat actors also try to adopt and deploy AI-enabled tools to

evade detection and compromise systems or data. Compliance with applicable privacy, cybersecurity, AI, data

protection, lawful access, content moderation and digital platform obligations can be costly and operationally

demanding and may require changes to our products, services, business practices, or technical infrastructure.

Environmental, Health, and Safety

Our operations and facilities, as well as existing and planned infrastructure, are subject to an extensive regulatory

framework of federal, state, local, and foreign environmental, health, and safety laws, and regulations and permits

that govern, among other things, employee health and safety, discharges of pollutants into the air and water, the

generation, handling, storage, and disposal of hazardous materials and wastes and the investigation and remediation

of certain materials, substances, and wastes. These include various regulations promulgated by federal, state, and

local regulatory agencies and legislative bodies. Certain of our operations, including launch, reentry, testing, and

manufacturing activities and the development or expansion of facilities, as well as the siting, construction and

operation of data centers, may require environmental reviews, consultations, and permits and may be subject to

conditions or mitigation measures that could increase costs or limit operations.

We are required to obtain a number of permits and entitlements from various government agencies to construct and

operate our facilities, including zoning, land use and building code permits, air quality permits for permanent

combustion equipment (including both diesel generators and natural gas turbines), stormwater and wastewater

discharge permits, and fire and life safety approvals. We have issued or pending permit applications for certain of

our facilities. For additional information, please refer to “Risk Factors—Risks Related to Our Business—

Environmental laws, regulations, litigation, liabilities and proceedings may adversely affect our operations,

225

including our launch operations, manufacturing activities, fuel storage and handling operations, launch facilities and

ground infrastructure, and data center operations and expansion plans.”

Government Contracts

A portion of our revenue is derived from contracts, directly or indirectly, with the U.S. government. We have

numerous direct contracts with the U.S. government, primarily NASA, the Department of War, the General Services

Administration, and certain Intelligence Community agencies. These contracts focus mainly on launch services,

spacecraft development, and satellite deployment, and artificial intelligence products. We are almost always the

prime contractor on our government contracts, and we rarely use subcontractors. All of our launch contracts with

U.S. government agencies are firm fixed-price contracts with milestone-based payments.

These contracts are subject to U.S. government contracting rules and regulations (Federal Acquisition Regulation

(FAR) and Defense Federal Acquisition Regulation Supplement (DFARS)), and therefore, we are subject to the

business risks specific to the defense industry. These regulations impose stringent requirements on our operations,

business practices and reporting, and noncompliance could result in civil or criminal penalties, suspension or

debarment from government contracting, or loss of existing or future business. These requirements, although

customary in U.S. government contracts, increase our performance and compliance costs. These costs might increase

in the future. The U.S. government has the ability to unilaterally: (i) declare us ineligible to receive new contracts;

(ii) terminate existing contracts at its convenience and without advance notice; (iii) reduce the scope and value of

existing contracts; (iv) audit our contract-related costs and fees, including allocated indirect costs; and (v) revoke

required security clearances. Violations of government procurement laws could result in civil or criminal penalties.

We are also required to maintain special security clearances and comply with executive orders, federal laws and

regulations, and customer security requirements for classified programs, and our government contracts impose

cybersecurity and information assurance requirements, including implementation of information security protections

in accordance with NIST Special Publication 800-171 and obligations to review and report certain cyber incidents.

Failure to comply could result in suspension of payments, termination of contracts, civil or criminal penalties, or

exclusion from future government contracting opportunities. For additional information, please refer to “Risk

Factors—Risks Related to Our Business—Our services are subject to risks related to supplying services to the U.S.

government.”

In addition, in connection with preparing leased real property for our launch operations at Kennedy Space Center

and Cape Canaveral, Florida, and Space Launch Complex 4 at Vandenberg Space Force Base, California, we make

significant capital improvements and install extensive real and personal property at these government-owned sites.

The launch facilities we build are a unique capital improvement compared to standard commercial use sites because

the federal government specifically designates these launch sites for aerospace activities, such as rocket launches.

Given the specific use requirements of these government-owned sites, we have historically entered into handover

agreements with the relevant government entities upon expiration or termination of the leases, pursuant to which the

improvements are transferred to the government rather than removed. This fact pattern has historically been the case

with previous leases such as at Cape Canaveral Space Force Station.

Legal Proceedings

We are involved in the legal proceedings described in Note 17, Commitments and Contingencies, in our audited

consolidated financial statements and Note 16, Commitments and Contingencies in our unaudited consolidated

financial statements included elsewhere in this prospectus, and we are subject to other claims and litigation arising in

the ordinary course of business. The outcome of any litigation is inherently uncertain, and if decided adversely to us,

or if we determine that settlement of particular litigation is appropriate, we may be subject to liability that could

have a material adverse effect on our business.

226

MANAGEMENT

Below is certain information as of May 1, 2026 regarding individuals who are expected to serve as our executive

officers and directors upon the completion of this offering.

Name	Age	Position
Elon Musk......................	54	Chief Executive Officer, Chief Technical Officer and Chairman of the Board
Gwynne Shotwell...........	62	President, Chief Operating Officer and Director
Bret Johnsen...................	57	Chief Financial Officer
Ira Ehrenpreis.................	57	Director
Randy Glein....................	60	Director
Antonio J. Gracias..........	55	Director
Donald Harrison.............	54	Director
Steve Jurvetson...............	59	Director
Luke Nosek.....................	50	Director

Executive Officers and Management Directors

Elon Musk has served as our Chief Executive Officer, Chief Technical Officer and Chairman of our board since

May 2002. Mr. Musk is also the Technoking of Tesla and has served as Chief Executive Officer of Tesla since

October 2008. Mr. Musk was Chief Technology Officer and on the board of directors of X, beginning October 2022

and served as the Chief Executive Officer and on the board of directors of xAI, beginning March 2023, in each case

through the March 2025 merger of X and xAI. Following the merger, Mr. Musk served as the President, Treasurer,

and Chief Executive Officer and on the board of directors of xAI, until it was acquired by the Company in February

2026. Mr. Musk is also a founder and Chief Executive Officer of Neuralink Corp., a company focused on

developing brain-machine interfaces, and The Boring Company, an infrastructure company. Prior to the Company,

Mr. Musk co-founded PayPal, an electronic payment system, which was acquired by eBay in October 2002, and

Zip2 Corporation, a provider of Internet enterprise software and services, which was acquired by Compaq in March

1999. Mr. Musk serves on the board of directors of Tesla and previously served on the board of directors of

Endeavor Group Holdings, Inc. from April 2021 to June 2022. Mr. Musk holds a B.A. in Physics from the

University of Pennsylvania and a B.S. in Business from the Wharton School of the University of Pennsylvania. Mr.

Musk brings to our board historical knowledge, operational and technical expertise, and continuity.

Gwynne Shotwell has served as our President and Chief Operating Officer since 2008 and has been a member of our

board since March 2009. Previously, Ms. Shotwell served as our Vice President, Business Development, from 2002

to 2008. Prior to joining the Company, Ms. Shotwell held positions with Microcosm, Inc., an aerospace company, as

a director, and The Aerospace Corporation, an independent, non-profit organization performing objective technical

analyses and assessments for a variety of government, civil, and commercial customers, as a senior project engineer.

Ms. Shotwell also serves on the board of directors of Polaris, Inc., a manufacturer of powersports vehicles, and on

Northwestern University’s Board of Trustees. Ms. Shotwell was inducted into the National Academy of Engineering

and was previously named the Satellite Executive of the Year, included on Time’s 100 Most Influential People, and

Fortune Magazine’s World’s 50 Greatest Leaders. Ms. Shotwell holds a B.S. in Mechanical Engineering and an

M.S. in Applied Mathematics from Northwestern University. As one of the key members of our leadership team,

Ms. Shotwell brings to our board extensive operational experience and in-house knowledge of the Company’s

operations, technology, research and development and business management.

Bret Johnsen has served as our Chief Financial Officer since 2011. In this role, Mr. Johnsen leads our global

finance organization and is responsible for our long-term financial strategy, internal financial operations,

interactions with the financial community, and the financial aspects of our growth initiatives. With more than two

decades of experience in financial leadership, primarily in high-profile technology and semiconductor companies,

his leadership continues to play a key role in driving our financial performance, long-term value creation and

operational discipline. Prior to joining the Company, Mr. Johnsen served as Chief Financial Officer at Mindspeed

Technologies, Inc., a publicly traded semiconductor company, from 2008 to 2011. Prior to that role, he spent nearly

227

a decade at Broadcom Inc., a global semiconductor company, from 1999 to 2008, holding roles of increasing

responsibility within the organization, including serving as Vice President and Corporate Controller. Mr. Johnsen

serves as a Trustee of the University of Southern California and holds a B.S. in Accounting from the University of

Southern California and an M.S. in Finance from San Diego State University, and he is a Certified Public

Accountant (CPA).

Non-Management Directors

Ira Ehrenpreis has served on our board since February 2026. Mr. Ehrenpreis is a founder and managing member of

DBL Partners, a leading impact investing venture capital firm, formed in 2015. Previously, he was a partner at

Technology Partners, a venture capital firm. Mr. Ehrenpreis serves on the board of directors of Tesla. He serves as

the Chairman of the VCNetwork, the largest and most active California venture capital organization. Mr. Ehrenpreis

also serves as the Chair of the National Association of Corporate Directors (NACD) Northern California and the Co-

Chair of the Stanford Precourt Institute for Energy Advisory Council. Among several other awards and honors, Mr.

Ehrenpreis has been named a member of the NACD Directorship 100 for being “one of the most influential leaders

in the boardroom and corporate governance community.” Mr. Ehrenpreis holds a B.A. from the University of

California, Los Angeles and a J.D. and M.B.A. from Stanford University. Mr. Ehrenpreis brings to our board

experience in the technology, impact and venture capital industries, as well as valuable insights in corporate

governance, strategic growth and shareholder values.

Randy Glein has served on our board since February 2026 and previously served as a board observer since 2009.

Mr. Glein is co-founder and managing partner of DFJ Growth, a venture capital firm that has invested in more than

100 growth-stage technology companies over the past 20 years. He currently serves on the board of directors of

several private technology companies and has previously served on the board of directors of Anaplan, Inc. and

Tremor Video, Inc. Prior to DFJ Growth, Mr. Glein served as Chief Financial Officer of FeedBurner (acquired by

Google in 2007) and Vice President of Tribune Company and its corporate investment group, Tribune Ventures. Mr.

Glein began his career in the aerospace industry as a systems engineer with Hughes Space & Communications and

in business development roles with its DIRECTV and New Ventures units. Mr. Glein holds a B.S.E.E. in Electrical

Engineering from the University of Florida, an M.S.E.E. in Electrical Engineering from the University of Southern

California, and an M.B.A. from the UCLA Anderson School of Management. Mr. Glein brings to our board

experience in the venture capital industry and more than 35 years of business and leadership experience in the

technology, media, and satellite communications industries.

Antonio J. Gracias has served on our board since October 2010. Since 2001, Mr. Gracias has been Chief Executive

Officer and Chief Investment Officer of Valor Management LLC, a private equity firm. As Founder, CEO, and CIO

of Valor, he oversees one of the leading growth-focused investment firms in the United States with over $55 billion

in assets under management. He has served on the board of Neuralink Corp., a company focused on developing

brain-machine interfaces, since May 2026, served on the board of The Boring Company, an infrastructure company,

since May 2026 and served as a director of Harmony Biosciences Holdings, Inc., a pharmaceutical company, from

September 2017 to May 2026. He also served as a director of Marathon Pharmaceuticals, LLC from November 2013

until its acquisition by PTC Therapeutics in May 2017, and SolarCity Corporation from 2012 to 2016. Mr. Gracias

previously served as a director of Tesla from 2007 to 2021 helping take the company public and acting as Lead

Independent Director for eight years. Prior to founding Valor Management LLC in 2001, Mr. Gracias served as

Founder and Managing Member of MG Capital, a private equity firm headquartered in Chicago, where he was the

lead transaction principal from 1995 through 2000. Prior to MG Capital, Mr. Gracias was an associate with

Goldman, Sachs & Co. in New York, where he served the firm’s institutional clients in the International Equity

Division. Mr. Gracias is also actively involved in philanthropic activities. He is a trustee of The Aspen Institute,

where he was a 2009 Henry Crown Fellow, an Aspen Institute program designed to engage the next generation of

leaders in the challenge of community-spirited leadership. Additionally, he serves as a member of several

prestigious non-profit and endowment boards, including the Board of Visitors for the Georgetown University School

of Foreign Service and the Pritzker School of Molecular Engineering at the University of Chicago. He is also a

member of the University of Chicago Board of Trustees. Mr. Gracias holds a joint B.S. and M.S.F.S. (Honors

Degree) in International Finance and Economics from the Georgetown University School of Foreign Service and a

J.D. from the University of Chicago Law School. Mr. Gracias brings to our board skills and experience in

228

investment strategy, portfolio company management and improvement, operations of business, and finance across

several industries, including aerospace, technology, and manufacturing.

Donald Harrison has served on our board since February 2015. Mr. Harrison has served as President, Global

Partnerships and Corporate Development at Google LLC, a technology company, since 2017. Mr. Harrison

previously served as Vice-President, Corporate Development at Google from 2012 to 2017 and as Vice-President

and Deputy General Counsel from 2005 to 2012. Mr. Harrison also sits on the board of directors of Reliance Jio, the

largest mobile telecommunications services provider in India. Mr. Harrison holds a B.A. in Philosophy and Political

Science from the University of King’s College and a J.D. and LLB from the Univ