Recursion Pharmaceuticals (RXRX) Q2 2026 Earnings Report

The call presents a strong set of operational and technical highlights: multiple advancing clinical programs with compelling early efficacy (REC-4881), a cleared IND for a precision-selective oncology candidate (REC-7735), large proprietary data assets (>50 PB, massive cellular atlases), meaningful partnership cash inflows (> $500M realized, > $260M from Roche-Genentech) and demonstrated efficiency gains in design speed and operating costs (guidance lowered to $375M, ~40% reduction vs 2024 pro forma). Key risks and lowlights are typical for a clinical-stage biotech: regulatory and clinical uncertainty in rare disease and CNS programs, timing of pivotal data (some milestones extend beyond the current cash runway), competitive pressure as AI tools proliferate, and the early-stage nature of some proofs. Overall, positive operational momentum and external validation from partners outweigh the remaining clinical and market risks, but the company still faces execution and timing risks as programs advance.

Management lowered 2026 full‑year cash operating expense guidance to $375 million (about a 40% reduction versus comparable 2024 pro forma expenses), reported cash and equivalents of approximately $557 million (runway through early 2028), and noted >$500 million in partnership cash inflows (including >$260 million with Roche‑Genentech and the opportunity for >$300 million of additional milestones per future small‑molecule program); they also cited engine and program metrics including >50 petabytes of multimodal data, ~1 trillion iPSC‑derived neurons and hundreds of billions of microglia, models comparing >17,000 genes across tens of millions of data points, AI agents driving 1.3–1.6x enrollment improvement, and Nesso‑1 delivering Boltz‑2 level accuracy with 10–20x faster inference; pipeline metrics included five clinical‑stage programs, REC‑4881 showing a median 43% polyp burden reduction at 3 months in Phase II, REC‑7735 designed in 10 months (242 compounds, 13 cycles) with >100‑fold selectivity for PIK3CA H1047R (IND cleared, Phase I to start later this year with dose‑escalation data expected H1 2028), and a design efficiency of ~330 compounds to candidate in ~1.5 years versus industry ~2,500 compounds over 4 years, with guidance to expect additional REC‑4881 Phase II data and a year‑end regulatory update, a fuller REC‑1245 update later this year, and REC‑7735 initiation.