JANUARY 2026 NAJAT KHAN, PHD CEO & PRESIDENT, RECURSION Recursion JP Morgan Presentation NK 25-Dec-25: Would move supercomputer image here NK 25-Dec-25: And a patient image – lets redo the design to make it fresh! Maintaining ambition while raising the bar NK 29-Dec-25: NEW – NK COMMENTS

This presentation of Recursion Pharmaceuticals, Inc. (“Recursion,” “we,” “us,” or “our”) and any accompanying discussion contain statements that are not historical facts may be considered forward-looking statements under federal securities laws and may be identified by words such as “anticipates,” “believes,” “estimates,” “expects,” “intends,” “plans,” “potential,” “predicts,” “projects,” “seeks,” “should,” “will,” or words of similar meaning and include, but are not limited to, statements regarding our ability to use AI to translate complex science into medicines faster and better; the amount and timing of potential milestone payments, cash position and cash runway; Recursion’s OS industrializing first- and best-in-class drug discovery; our ability to industrialize clinical development and the effect of doing so on clinical trial outcomes; the occurrence or realization of potential milestones; current and future preclinical and clinical studies, including timelines for enrollment in studies, data readouts, progression toward IND-enabling and other potential studies, and engagement with the FDA; advancements of and other decisions regarding our pipeline, partnerships, and data strategies; the potential size of the market opportunity for our drug candidates; outcomes and benefits from licenses, partnerships and collaborations, including option exercises by partners; the initiation, timing, progress, results, and cost of our research and development programs; advancements of our Recursion OS; the potential for additional partnerships; our ability to identify viable new drug candidates for clinical development and the accelerating rate at which we expect to identify such candidates including our ability to leverage the datasets acquired through the license agreement into increased machine learning capabilities and accelerate clinical trial enrollment; and many others. Other important factors and information are contained in Recursion’s most recent Annual Report on Form 10-K, Quarterly Report on Form 10-Q, and the Company’s other filings with the U.S. Securities and Exchange Commission (the “SEC”), which can be accessed at https://ir.recursion.com, or www.sec.gov. All forward-looking statements are qualified by these cautionary statements and apply only as of the date they are made. Recursion does not undertake any obligation to update any forward-looking statement, whether as a result of new information, future events or otherwise. Certain information contained in this presentation relates to or is based on studies, publications, surveys and other data obtained from third-party sources and the company’s own internal estimates and research. While the company believes these third-party sources to be reliable as of the date of this presentation, it has not independently verified, and makes no representation as to the adequacy, fairness, accuracy or completeness of, any information obtained from third- party sources. In addition, all of the market data included in this presentation involves a number of assumptions and limitations, and there can be no guarantee as to the accuracy or reliability of such assumptions. Finally, while the company believes its own internal research is reliable, such research has not been verified by any independent source. Information contained in, or that can be accessed through our website is not a part of and is not incorporated into this presentation. Cross-trial or cross-candidate comparisons against other clinical trials and other drug candidates are not based on head-to-head studies and are presented for informational purposes; comparisons are based on publicly available information for other clinical trials and other drug candidates. Any non-Recursion logos or trademarks included herein are the property of the owners thereof and are used for reference purposes only. Important Information 2

OUR MISSION: Decoding biology to radically improve lives

WHAT WE BUILT: A unified, AI-native intelligence platform that translates complex science into medicines that matter — faster, better, and at scale for the patients who are waiting

LEADING THE EVOLUTION OF HOW MEDICINES ARE DISCOVERED AND DELIVERED Novel medicines that matter The first AI-native end-to-end platform, from idea through the clinic Precision Design Novel Biological Discoveries Next-Gen Clinical Development Bilingual teams & culture: fluent in science and tech Purpose-built models & integrated compute Proprietary, multimodal data at unprecedented scale

1. Includes preclinical programs that are expected to enter the clinic within the next 18 months 2. Potential Roche and Genentech and Sanofi milestones per small molecule program 3. Cash, cash equivalents and restricted cash (unaudited) as of December 31, 2025 4. Year-end 2027 runway guidance includes risk-adjusted cash inflows from partnerships Recursion: Progress, by the numbers Wholly owned1~5 CLINICAL DEVELOPMENT ✓FAP: FIRST AI-enabled clinical proof-of-concept with potential first-in-class profile ✓>$500M in upfront & milestones achieved • >$300M in potential milestones + up to double digit royalties per program2 • Several high value opportunities Partnered Wholly owned ~15 DISCOVERY The first AI-native end-to-end platform, from idea through the clinic $755 million in YE 2025 cash3, providing expected runway through YE 20274 6

How we will create impact – with focus and discipline 7 Translate insights ➔ proof points ➔ new medicines Focused innovation, grounded in clear impact Empower exceptional, bilingual teams to deliver impact and humanity Pair bold ambition with disciplined execution 1 2 3

8 Ongoing momentum Where we’re going First clinical POC for Recursion’s platform • REC-4881 – meaningful & durable polyp reduction with safety profile consistent with class Advance clinical validation • REC-4881 – optimize dosing schedule & FDA-aligned registrational study plan • ~5 programs entering or in trials with upcoming go/no-go decisions Deliver differentiated programs with partners • Advance programs to late-discovery value inflection points • Translate biology maps into new, partner accepted discovery programs External validation of Recursion’s platform • Sanofi: 4 milestones achieved for AI-designed molecules • Roche and Genentech: 6 AI-powered biology maps delivered 1. Translate insights to proof points – on the path to new medicines

9 Ongoing momentum Where we’re going Unprecedented scale in phenomics • At-scale cellular imaging uncovering novel biology Mature, actionable portfolio of high-quality targets • Integrate omics and patient data with purpose-built models (e.g. transcriptional Foundational Models) 2. Focused innovation, grounded in clear impact Generative drug design with predictive precision • AI-multi-parameter optimization paired with structure- based systems AI-powered Clinical Development – initiated • Natural history (e.g. FAP) and biomarker (e.g. CDK7) analysis • AI-driven site selection and recruitment Generative drug design at scale • Next-gen models and agentic systems for design • Design drives differentiation & speed for programs AI-powered Clinical Development at scale • Deployed across programs and stages • Increased automation and in silico trial design and execution

10 Ongoing momentum Where we’re going Disciplined prioritization of work and spend • Portfolio focused to ~5 clinical programs (May 2025) • Streamlined operations by 35% Rapid decisions to ensure strategic capital allocation • Portfolio decisions anchored in automated Target Product Profiles • Tooling and automation for objective decision-making • Disciplined platform investment tied to impact 3. Pair bold ambition with disciplined execution Drive enterprise efficiency • Deploy AI-agents to reduce cost, increase speed, or add flexibility Expected 2026 cash burn1 of <$390 million ~35% reduction in pro forma operating expenses 2024 to 20262 1. Cash burn—defined as operating cash flow less capital expenditures, excluding partnership and financing inflows, transaction expenses—is a non-GAAP financial measure. See Appendix for reconciliation of non-GAAP financial measures. 2. YE2024 reported OpEx for Recursion and Exscientia combined, excluding non-cash GAAP items (e.g. share-based compensation). 2026 estimate of <$390 million cash burn One Recursion mindset post integration • Unified teams, platform, and operating model

Wholly owned clinical pipeline Translate insights ➔ proof points ➔ new medicines

Wholly owned pipeline: Translating insight into proof Differentiation powered by the Recursion OS — from biology to design to clinical development 12 Target Disease Indication Late Discovery Preclinical Phase 1/2 Phase 3 Potential Milestone REC-4881 MEK1/2 Familial adenomatous polyposis (FAP) FDA engagement – 1H26 REC-617 CDK7 Advanced solid tumors Combination data – 1H27 REC-1245 RBM39 Biomarker-enriched solid tumors & lymphoma Ph 1/2 data – 1H26 REC-3565 MALT1 B-cell malignancies Ph 1 data – 1H27 REC-4539 LSD1 Solid tumors & hematology oncology Ph 1 trial start – 1H26 REC-7735 PI3Kα H1047R HR+ breast cancer Go/no-go decision – 2H261 REC-102 ENPP1 Hypophosphatasia (HPP) Go/no-go decision – 2H261 1. Data-driven decision for potential Phase 1 initiation

An AI-native, end-to-end platform for drug discovery and clinical development 13 In Silico Binding Affinity Automated Testing Synthesis Aware Design Physics-based Models ML Property Prediction Transcriptomics Patient Data (RWD) LLMs & Graphs Target ValidationPhenomics Clinical Trial Design AI-powered Recruitment Causal AI Patient Selection & RWD Insight to MoleculeBiology to Insight Molecule to Patient Generative Chemistry Automation Automation Supported by BioHive-2 The fastest, most powerful supercomputer wholly-owned and operated by a pharma or biotech company ~65PB of data including 40PB of proprietary data Automation

How the platform compounds value across programs MEK1/2 RBM39 CDK7 MALT1 LSD1 ENPP1 PI3K⍺ H1047R Late Discovery Early Discovery Partnered Discovery Phenomics Transcriptomics Patient Connectivity & Other Molecular Design 3D Protein & Atomistic Models Automated Chem, Bio, & ADMET Causal AI & RWD AI-powered Recruitment Trial Design I n s ig h t to M o le c u le B io lo g y t o I n s ig h t M o le c u le t o P a ti e n t Illustrative V2.0V1.0V0.1 To be utilized as programs advance and at partner discretion 14

FAP: Significant unmet need characterized by lifelong polyp progression in the GI tract, with no approved pharmacotherapies 15 1. Internal company estimates 2. 20-30% of FAP patients have de novo mutations with no family history. Half E et al, Orphanet J Rar Dis. 2009 Colectomy: Colon removal Removal of rectum & high-risk polyps Duodenectomy: Duodenum removal Lifelong continuum of disease progression and intervention, driven by chronic polyp growth • Surgery has no impact on slowing disease progression • Lifetime of endoscopic surveillance, frequent excisional interventions, life-altering surgeries, and poor QoL, morbidity and mortality Disease progression & severity REC-4881 may be positioned to fill a significant unmet need with no approved pharmacotherapies • Orphan disease caused by autosomal dominant inactivating mutations in APC2 • Near-100% CRC risk for patients by age ~40 in absence of surgery or excisions • Adenomas progressively accumulate with limited evidence of spontaneous regression Familial adenomatous polyposis (FAP) affects >50,000 patients in US + EU51 Polyps on mucosal membrane of colon Colonoscope Scope view

Platform insight: Unbiased AI-driven phenotypic discovery identified MEK1/2 inhibition as a novel APC-rescue mechanism 16 Guided by this novel insight, Recursion in-licensed REC-4881 from a major pharmaceutical company1 and redirected REC-4881 as a mechanistically aligned therapeutic candidate for FAP 1. Recursion in-licensed REC-4881 from Takeda REC-4881 Similar Opposite • Recursion’s unbiased AI- driven phenotypic approach uncovered MEK1/2i as a mechanism that rescues APC-deficient cells • Platform screened thousands of compounds in APC mutant rescue assays - REC-4881 strongly clusters with APC rescue signature Phenotypic discovery identified a novel, therapeutic entry point for FAP Natural history study confirms high unmet need and FAP disease progression when untreated

17 FAP registry analysis showed: • 87% of untreated FAP patients had annual polyp- burden increase • Average of 60% polyp burden increase annually Note: Natural History Analysis evaluated data from ~200 FAP patients; analysis shown represents a subset of patients who satisfy key inclusion criteria of TUPELO. Study is intended to Contextualize the TUPELO single arm data, better understand background natural history of FAP disease progression. Study limi tations include potential variability in polyp count estimation between endoscopies; endoscopies are typically conducted annually in routine care, while the TUPELO data represent polyp burden at Week 13 1. Notes were processed using a custom NLP pipeline with Recursion-built and operated supercomputer, BioHive-2 2. Major surgeries were defined as ileal pouch-anal anastomosis, colectomy, proctocolectomy, Whipple, i leostomy, and ampullectomy. Analysis captures documentation of “history of” such procedures. Major surgeries are expected to be underestimated due to limited follow up period and potential underreporting of patient medical history. LLM analysis of 256,000 U.S. physician notes1 of FAP patients: Relentless disease progression and continuous intervention. 75% of patients had a major FAP-related surgery documented2 Annualized % change in polyp burden in a natural history cohort Amsterdam University Medical Center FAP registry (N=55) Platform insight: Natural history analysis revealed ~87% of patients have polyp burden increase, highlighting the disease progression of FAP Phenotypic discovery identified a novel, therapeutic entry point for FAP Natural history study confirms high unmet need and FAP disease progression when untreated “No polyps were removed given there were an innumerable amount (>300) and removal would not be endoscopically feasible." “[they] had greater than 10 surgeries and multiple revisions of [their] colectomy and resection of [their] small bowel.”

18 Safety & efficacy: Rapid reductions in polyp burden with 4mg dose of REC-4881 and safety profile consistent with class effects Note: Polyp burden defined as the sum of all diameters of polyps in the GI 1. Following the March data cut, a quality review identified suboptimal bowel preparation at baseline. To ensure an accurate, like-for-like assessment, polyp burden was re- evaluated using video review restricted to the clean distal LGI segments matched to the same anatomical regions at Weeks 13 and 25 2. Patient reached W25 but did not perform W25 Assessment 3. Efficacy Evaluable Population (n=12): Defined as all participants who have measurable disease (non-zero polyp burden) at end of baseline endoscopy, received at least 75% of study drug, and have at least one post-baseline on study endoscopic assessment. One patient was efficacy evaluable after completion of W25 assessment but did not complete W13 assessment, baseline measurement carried forward for W13 assessment per SAP for missing data. Therefore, this patient contributed 0% polyp burden reduction at W13 and not shown in figure. 4. Discontinuations: Grade 1 (n=1): 1 diarrhea, Grade 2 (n=3): 1 retinopathy, 1 rash, 1 hypertension Note: N of 12 patients were efficacy evaluable, 1 patient missed Week 13 assessment Data Cutoff Date: 2025-11-25 Percent (%) change from baseline calculates the change between post-resection value from screening visit to the pre-resection value at Week 13/EOT visit. Subjects with absolute value of 0 at baseline are not displayed. REC-4881 4mg % change from baseline in total polyp burden at Week 13/EOT (12 weeks on therapy) % c h a n g e f ro m b a se li n e 1 2 Safety profile consistent with MEK1/2 inhibition • 18 TRAE events with majority Grade 1/2: • E.g., Dermatitis acneiform, CPK increases, rash, diarrhea, LVEF decrease • Low rates of Grade 3 TRAEs (n=3) • No Grade 4/5 events • Discontinuations (n=4)4 75% evaluable patients responded • Polyp burden reduction3: 43% median Summary of Adverse Events On Treatment Phase | Week 13

19 Durability: Durable reductions in polyp burden maintained with 4mg dose of REC-4881 Note: Polyp burden defined as the sum of all diameters of polyps in the GI 1. Following the March data cut, a quality review identified suboptimal bowel preparation at baseline. To ensure an accurate, like-for-like assessment, polyp burden was re-evaluated using video review restricted to the clean distal LGI segments matched to the same anatomical regions at Weeks 13 and 25 2. Patient reached W25 but did not perform W25 Assessment 3. Efficacy Evaluable Population (n=12): Defined as all participants who have measurable disease (non-zero polyp burden) at end of baseline endoscopy, received at least 75% of study drug, and have at least one post-baseline on study endoscopic assessment. One patient was efficacy evaluable after completion of W25 assessment but did not complete W13 assessment, baseline measurement carried forward for W13 assessment per SAP for missing data. Therefore, this patient contributed 0% polyp burden reduction at W13 and not shown in figure. Data Cutoff Date: 2025-11-25 • 82% of evaluable patients responded • 73% achieved durable ≥30% reductions • Polyp burden reduction: 53% median REC-4881 4mg dose level % change from baseline in total polyp burden at Week 25 (12 weeks off therapy) % c h a n g e f ro m b a s e li n e Percent (%) change from baseline calculates the change between post-resection value from screening visit to the pre-resection value at Week 25/EOT visit. 1 Off Treatment Phase | Week 25

REC-4881 (MEK1/2): First clinical validation of Recursion’s platform – disease with no approved pharmacotherapies Biological insight: MEK1/2 inhibition identified as a unique mechanism for FAP Clinical proof: Rapid and durable polyp burden reduction with adverse events consistent with MEK inhibitors What’s next: Define registration path; continue dose optimization REC-4881 Target Profile • Addressing high unmet need for >50,000 addressable patients with no approved pharmacotherapies • Orally bioavailable, highly potent and selective MEK1/2 inhibitor • Differentiated ADME profile may enhance exposures at the site of GI adenomas • ODD in US and EU; FTD in US 20 Insight Proof points New medicines

Translate insights ➔ proof points ➔ new medicines Partnered Discovery

Power of Recursion OS in advancing partnered drug discovery Partnered Discovery Phenomics Transcriptomics Patient Connectivity & Other Molecular Design 3D Protein & Atomistic Models Automated Chem, Bio, & ADMET Causal AI & RWD AI-powered Recruitment Trial Design I n s ig h t to M o le c u le B io lo g y t o I n s ig h t M o le c u le t o P a ti e n t Illustrative Discovering Novel Biology Precision Design Targeted Patient in total cash inflows achieved across all our partnerships and collaborations >$500 million 22 Select progress-based milestones achieved: Roche and Genentech ($30m) – Neuron map – 3Q24 Sanofi ($7m) – I&I 3 – 2Q25 Roche and Genentech ($30m) – Microglia map – 4Q25 Sanofi ($4m) – I&I 1 – 3Q23 Sanofi ($11m) – I&I 2 – 3Q24 Sanofi ($4m) – Oncology 1 – 3Q24 Potential for >$300m per small-molecule program

Roche and Genentech: Two groundbreaking maps in neuroscience provide a whole-genome view of brain’s biology 23 1 trillion human-induced pluripotent stem cell (hiPSC)-derived neuronal cells produced 100 billion+ microglial cells produced using new cell manufacturing techniques 2 first-of-its-kind neuroscience maps Proprietary foundation models powered by our supercomputer, BioHive-2 Disease-like perturbations to cells, including knockout and over-expression spanning tens of 1000s of genes Wet lab Dry lab

Roche and Genentech: Track record of delivery on biological maps and insights Phenomics Transcriptomics Patient Connectivity & Other Molecular Design 3D Protein & Atomistic Models Automated Chem, Bio, & ADMET Causal AI & RWD AI-powered Recruitment Trial Design I n s ig h t to M o le c u le B io lo g y t o I n s ig h t M o le c u le t o P a ti e n t Proprietary digital maps of complex biology and chemistry Collaboratively develop novel therapeutic programs directly from identified biological insights 6 AI powered- Phenomaps in neuroscience and 1 GI oncology indication 24 Next steps: Continue advancement of multiple neuroscience insights into target validation

Sanofi: Advancing differentiated, potential best-in-class molecules in oncology and I&I Phenomics Transcriptomics Patient Connectivity & Other Molecular Design 3D Protein & Atomistic Models Automated Chem, Bio, & ADMET Causal AI & RWD AI-powered Recruitment Trial Design I n s ig h t to M o le c u le B io lo g y t o I n s ig h t M o le c u le t o P a ti e n t 4 Program milestones achieved to date Next steps: Advance programs to lead optimization and development- candidate milestones Adaptable and scalable platform delivering novel chemical matter against difficult and diverse protein targets Active learning to overcome data poor project challenges and synthesize molecules efficiently 25

Recursion OS Platform Focused innovation, grounded in clear impact

Focused investment: Transcriptional Foundational model for actionable, novel target discovery 27 Biology to Insight Transcriptional Foundation Model Cross-model RNA data A scalable portfolio of high-quality targets grounded in patient biology Why: Transcriptional state provides a scalable, cross- system readout of disease biology — enabling lab-to- patient translation and actionable target discovery in silico ✓Best-in-class Transcriptomic Model: Delivers state- of-the-art biological representations for target discovery ✓Generalization across system and scales: Works across patient + in vitro data, bulk and single-cell RNA ✓Data efficient learning: Reaches comparable performance with ~50× less training data Insight to Molecule Molecule to Patient Translational Spectrum Unified Data Modalities patient & in vitro samples bulk & single cell RNAseq

• 100 million+ molecules generated using synthetically aware design • >95% AI-generated, scored, and prioritized – all patentable • ~330 compounds to an advanced candidate in ~17 months (on average)1 • >10 development candidates designed across programs Focused investment: AI-driven high quality generative design at scale 28 1. Compared to industry average of over 2,500 compounds and 42 months Repeatable, time-compressed impact — enabling disciplined portfolio advancement Insight to Molecule Biology to Insight Molecule to Patient

Focused investment: AI-driven clinical development for trial design and execution 29 Molecule to Patient Data Foundation: ~1M molecularly profiled lives + ~300M real-world lives Application 1: Causal AI-based human genetics • Target validation and patient selection across portfolio Application 2: Smart trial design & feasibility with impact • 10-40% increase in eligible population • ~1.5X improvement in enrollment rate • Site & country selection in hours vs. months Insight to MoleculeBiology to Insight

Looking ahead MM: Alt photo option

Expected upcoming milestones 2026 and 2027 pipeline and partnership catalysts 31 1. Data-driven decision for potential Phase 1 initiation 2. Cash, cash equivalents and restricted cash (unaudited) as of December 31, 2025 3. Year-end 2027 runway guidance includes risk-adjusted cash inflows from partnerships 4. Cash burn—defined as operating cash flow less capital expenditures, excluding partnership and financing inflows, transaction expenses—is a non-GAAP financial measure. See Appendix for reconciliation of non-GAAP financial measures. Translate insights ➔ proof points ➔ new medicines Pair bold ambition with disciplined execution Focused innovation, grounded in clear impact 1H 2026 REC-4881 (MEK1/2i) Engage with FDA REC-1245 (RBM39 degrader) Mono - early safety and PK 2H 2027 REC-4539 (LSD1i) Mono - Early safety and PK 1H 2027 REC-4881 (MEK1/2i) Additional clinical data REC-617 (CDK7i) Combo - early safety and PK REC-3565 (MALT1i) Mono - early safety and PK • $755M in YE 2025 cash2 with expected runway through YE 20273 • Expected 2026 cash burn4 of <$390 million • Biology foundation models and patient data enabling scalable, high- quality target discovery • Generative AI design at scale (next-gen models and agentic systems) • Clinical Development AI at scale Partner catalysts – 2026 & 2027 Later-stage discovery milestones Advancing maps to early-stage programs Anticipated multiple new project initiations 2H 2026 Go/no-go decision1 REC-7735 (PI3Kα H1047Ri) REC-102 (ENPP1i)

32 THANK YOU

Appendix

To supplement our financial statements prepared in accordance with U.S. GAAP, we monitor and consider cash burn, which is a non-GAAP financial measure. We define cash burn as the net cash used in operating activities, excluding non-ordinary course transaction costs, plus partnership cash inflows and purchases of property and equipment. This non-GAAP financial measure is not based on any standardized methodology prescribed by U.S. GAAP and is not necessarily comparable to similarly-titled measures presented by other companies. We believe cash burn to be a liquidity measure that provides useful information to management and investors about the amount of cash consumed by the operations of the business, including our purchases of property and equipment. A limitation of using this non-U.S. GAAP measure is that cash burn does not represent the total change in cash and cash equivalents for the period because it excludes cash provided by or used for other investing and financing activities. We account for this limitation by providing information about our capital expenditures and other investing and financing activities in the statements of cash flows in our financial statements and by presenting cash flows from investing and financing activities in our reconciliation of cash burn. In addition, it is important to note that other companies, including companies in our industry, may not use cash burn, may calculate cash burn in a different manner than we do or may use other financial measures to evaluate their performance, all of which could reduce the usefulness of cash burn as a comparative measure. Because of these limitations, cash burn should not be considered in isolation from, or as a substitute for, financial information prepared in accordance with U.S. GAAP. Non-GAAP Financial Measures 34