Slide 1: Earnings 1Q26
2) Recursion. Earnings 1Q26 MAY 6, 2026
Slide 2: Important Information
Important Information This presentation of Recursion Pharmaceuticals, Inc. (“Recursion,” “we,” “us,” or “our”) and any accompanying discussion cont ain statements that are not historical facts may be considered forward-looking statements under federal securities laws and may be identified by words s
Slide 3: Our journey: Turning Recursion's foundation into proof with clinical and partnered programs
Our journey: Turning Recursion’s foundation into proof with clinical and partnered programs INITIAL CLINICAL & AI POTENTIAL PHASE DISCOVERY ASSETS PARTNERED ASSETS Foundational Proprietary Data Lab in Clinical & Partnered Infrastructure Engine a Loop Advancement Automated lab infrastructure Building
Slide 4: Recursion today: Delivering proof with a scalable, repeatable product engine
Recursion today: Delivering proof with a scalable, repeatable product engine CLINICAL PROOF ESTABLIS HED MULTIPLE SHOTS ON GOA L PROVEN PARTNER MODEL REC-4881 for FAP 5 wholly-owned programs >$500M in partner inflows and demonstrates first clinical with defined inflection points 10+ milestones deliv
Slide 5: Our focus: Advancing internal and partnered pipeline
Our focus: Advancing internal and partnered pipeline by translating our AI-native product engine into therapeutic impact Proprietary Multimodal Data Integrated Wet/Dry Lab Wholly owned pipeline AI/ Foundation Differentiated Models medicines, powered by AI Super Compute Partnered pipeline Bilingual T
Slide 6: Three strategic pillars driving the transition from proof to products
Three strategic pillars driving the transition from proof to products 1 2 3 Scaling a Pairing bold ambition Translating proof differentiated AI- with disciplined to products native product engine execution Translate deep pipeline Advance a proprietary, Pairing focused innovation learnings into appro
Slide 7: Wholly owned pipeline: A differentiated, platform-derived pipeline
Wholly owned pipeline: A differentiated, platform-derived pipeline with clear catalysts and disciplined decision points Total Addressable Late Disease Indication(s) Market1 Discovery Preclinical Phase 1/2 Phase 3 Differentiation REC-4881 Familial adenomatous • Targets underlying APC biology; rapid,
Slide 8: Wholly owned pipeline: Recent progress and next steps
Wholly owned pipeline: A differentiated, platform-derived pipeline with clear catalysts and disciplined decision points Total Addressable Late Disease Indication(s) Market1 Discovery Preclinical Phase 1/2 Phase 3 Recent Progress & Next Steps REC-4881 Familial adenomatous >50,000 ✓ Regulatory engagem
Slide 9: Partnered pipeline: Delivering AI-firsts with partners
Partnered pipeline: Delivering AI-firsts with partners on a robust & diverse joint portfolio of programs Select progress-based milestones achieved: 12-month outlook: ACCEPTED ACCEPTED ACCEPTED ACCEPTED Sanofi: Potential Sanofi Sanofi Sanofi Sanofi I&I 1 program Onc 1 program I&I 3 program Onc 2 prog
Slide 10: Vicki Goodman, M.D. appointed as Recursion CMO
Vicki Goodman, M.D. appointed as Recursion CMO • Seasoned Physician Executive: More than two decades of experience in oncology drug development and medical leadership, across Merck, Bristol Myers Squibb, GlaxoSmithKline, Exelixis, and Mural Oncology • Strategic Clinical Leadership: Former CMO and EV
Slide 11: REC-1245
Translating proof to products REC-1245
Slide 12: RBM39 Degrader: Selective degradation of RBM39 impairs splicing
RBM39 Degrader: Selective degradation of RBM39 impairs splicing, compromising DDR pathways and transcriptional regulation Biological Rationale • RBM39 is a splicing factor essential for DDR network and a c-Jun/ER/PR transcriptional coactivator • Overexpression of RBM39 in many solid tumors is linked
Slide 13: REC-1245: Novel platform derived insight to unlocking genomic instability vulnerabilities
REC-1245: Novel platform derived insight to unlocking comprehensive genomic instability vulnerabilities Recursion OS Insight Preclinical Insight Preclinical Data In vitro: Cell viability with REC-12452 Ovarian CDX Model: OVK18 (MSI-H) CDK13 CDK12 RBM39 Tumor Volume (mm3) RBM39 CTG IC50 (µM) CDK12 CD
Slide 14: Phase 1 Dose Escalation Update: Early data from REC-1245
Phase 1 Dose Escalation Update: Early data from monotherapy dose escalation with REC-1245 Key inclusion criteria Ph 1A Monotherapy Dose-Escalation • Unresectable, locally recurrent, or metastatic select solid Continuous once-daily dosing summary tumors or select relapsed/refractory lymphoma Addition
Slide 15: Preliminary Safety Data: REC-1245 well-tolerated with no DLTs
Preliminary Safety Data: REC-1245 well-tolerated with no DLTs across all evaluated doses to date Treatment-Related Adverse Event (TRAE) Preliminary safety and tolerability summary Patients (n=16) • REC-1245 was well-tolerated • No DLTs reported across evaluated doses to date Patients with Any TRAE 1
Slide 16: Preliminary PK/PD summary
Preliminary PK/PD summary • Predictable, dose-dependent exposure across evaluated patients to date Early data suggests • PK is supportive of QD dosing and exposures continue to increase with dose REC-1245 has predictable, dose • Expect to achieve exposures consistent with dependent tumor regression
Slide 17: REC-1245 (RBM39): Insight → early proof → next steps
REC-1245 (RBM39): Insight → early proof → next steps Proof points Insight Next steps Early clinical proof: Biological insight: Well-tolerated with no What’s next: Platform-derived novel DLTs in early Phase 1 Continued Phase 1 target and degrader, dose escalation; dose escalation; enabling synthetic
Slide 18: REC-4539
Translating proof to products REC-4539
Slide 19: LSD1i: Promising oncology target blocked by class-limiting toxicity
LSD1i: Promising oncology target historically blocked by class- limiting on-target toxicity and poor CNS exposure • Overexpression of LSD1, a pivotal epigenetic master regulator, Challenges promotes tumor progression and immune evasion • Potential to address high-impact indications by targeting LSD1
Slide 20: REC-4539: AI-enabled precision design to overcome class-limiting toxicity
REC-4539: AI-enabled precision design to overcome class-limiting toxicity issues Recursion OS Insight Preclinical Insight Preclinical Data SCLC CDX Model: H1417 Development Candidate (DC) Criteria: • Brain:plasma ratio: green >0.5; red <0.5 • MDCK-MDR1 efflux ratio (Pgp): green <2; yellow >2-<10; re
Slide 21: FPD in Phase 1 Dose Escalation: ENLYGHT Phase 1 trial
FPD in Phase 1 Dose Escalation: ENLYGHT Phase 1 trial in patients with select solid tumors including SCLC Phase 1 Dose Escalation • Rapid data-driven program go/no-go ENLYGHT patient population; N = ~25 based on clinical safety profile observed in • Patients with solid tumors including SCLC and high
Slide 22: REC-4539 (LSD1): What's next
REC-4539 (LSD1): What’s next Proof points Insight New medicines Preclinical proof: Biological insight: What’s next: Well-tolerated with Shorter-predicted human minimal impact Early monotherapy on platelets and similar safety and PK half-life vs competitors plus preclinical efficacy to potentially re
Slide 23: Recursion OS Platform
Scaling a differentiated AI-native product engine Recursion OS Platform
Slide 24: AI-native product engine for drug discovery & development
AI-native product engine for drug discovery & development 30-60% faster enrollment for clinical trials 10-40% increase in eligible patient population ~90% fewer compounds synthesized2 ~2x faster candidate advancement 10+ high-dimensional disease maps1 Multiple novel target programs, more upcoming Ve
Slide 25: A scaled, closed-loop data engine for learning human biology
A scaled, closed-loop data engine for learning human biology Automated wet lab generates high-quality, perturbation-driven data to power predictive models Recursion is a leader in generating perturbational data at scale • High content imaging/phenomics and transcriptomics provide a scalable readout
Slide 26: TxPert: Predicting biology before we run the experiment
TxPert: Predicting biology before we run the experiment New model predicts transcriptomic readouts for unseen perturbations Published in Nature Biotechnology Learns complex biology, moving beyond simple pattern recognition TxPert predicts cellular response to • Grounded in biological knowledge graph
Slide 27: TxFM: A biological foundation model that connects lab and patient biology
TxFM: A biological foundation model that connects lab and patient biology State-of-the-art transcriptomics foundation model Technical Achievements Model Performance Presented at ICLR 2026 ✓ Removes experimental noise TxFM outperforms 16 other foundation models and baselines TxFM bridges the gap betw
Slide 28: Financials
Pairing bold ambition with disciplined execution Financials
Slide 29: Cash runway to deliver on upcoming milestones
Cash runway to deliver on upcoming milestones 1Q26 expense and cash1 update • ~30% reduction in YoY cash opex 2 to $85m in 1Q26 • $665 million in cash1 as of March 31, 2026 1Q26 partnership highlights • 5th Sanofi milestone: $4M for a lead series in a potential first-in-class oncology program Reiter
Slide 30: Looking ahead
Looking ahead
Slide 31: Expected upcoming milestones
Expected upcoming milestones 2026 and 2027 pipeline and partnership catalysts 1H 2026 1H 2027 REC-4881 (MEK1/2i) REC-4881 (MEK1/2i) Initial engagement with FDA Additional clinical data REC-1245 (RBM39 degrader) REC-617 (CDK7i) Mono - early safety and PK Combo - early safety and PK REC-3565 (MALT1i)
Slide 32: Thank You
é ) . re 4 aw ay ; we 54", ee - = Bees ¢ - 2 ; F al * (ann ; er" a t > ‘a rr? | ¥ > * i 4 nin NY is ~ ’ vaca’ € 1
Slide 33: Slide 33
Slide 34: Appendix: Non-GAAP Financial Measures
Appendix Non-GAAP Financial Measures To supplement our financial statements prepared in accordance with U.S. GAAP, we monitor and consider operating cash burn, which is a non-GAAP financial measure. We define operating cash expense as the net cash used in operating activities, excluding non-ordinary