Recursion Pharmaceuticals

Recursion Pharmaceuticals is an AI-native ("TechBio") drug-discovery platform, not a single-drug biotech. It combines a proprietary phenomics data flywheel (petabytes of high-content imaging from millions of automated wet-lab experiments), the BioHive-2 NVIDIA-GPU supercomputer, and AI foundation models (Phenom, in NVIDIA's BioNeMo ecosystem) to industrialize target discovery, hit-to-lead, and lead optimization. The 2024 acquisition of Exscientia added generative chemistry + automated synthesis. Lead program REC-4881 is a repurposed MEK1/2 inhibitor for familial adenomatous polyposis (FAP); REC-1245 (RBM39 degrader) and REC-7735 (mutant-selective PI3Kα) are platform-origin oncology assets.

• monetization model: Collaboration/license revenue from pharma partners — upfronts, research funding, milestones, royalties; no product sales. (10-K Item 1; Note 3. sec.gov ; acc. 0001601830-24-000014)
• revenue mix: FY2023 total revenue ~$44.6M; 100% collaboration. Bayer ~50%, Roche/Genentech ~43%, Takeda ~7%; no product revenue. (10-K Item 1; Note 3. same URL; acc. 0001601830-24-000014)
• who pays: Pharma/biotech R&D partners: Bayer, Roche/Genentech, Takeda. (10-K Item 1; Note 3. same URL; acc. 0001601830-24-000014)
• margin_value_capture: No product gross margin; operating losses (FY2023 net loss ~$328M on ~$44.6M revenue). Value accrues via partner milestones/royalties if pipeline succeeds. (10-K Item 7/8. same URL; acc. 0001601830-24-000014)

• customer classes: Pharma & Biotech

Recursion does not report multiple reportable segments in the way a typical software or product company does. The revenue model is contract- and milestone-based, not unit-based product sales.

• Product revenue: Recursion has no approved drug product and therefore reports no product sales revenue. The 10-K for fiscal 2023 describes the company as a clinical-stage biopharmaceutical company that has not generated revenue from product sales and does not expect to do so for the foreseeable future.

Source: Recursion Pharmaceuticals, Inc. 10-K FY2023 — EDGAR issuer page: sec.gov ; accessed 2026-04-26; qualitative claim from filing; exact accession/exhibit not verified in this session.

• Collaboration and service revenue: This is effectively 100% of reported revenue. Collaborations with Roche/Genentech and Bayer are the named material agreements.

Source: Recursion Pharmaceuticals, Inc. 10-K FY2023 — business/risk disclosures; EDGAR URL above; exact accession/exhibit not verified in this session.

• Segment mix: Because the company is effectively a single-segment, clinical-stage company with collaboration revenue, a revenue-mix table like “software vs. services vs. product” does not apply. The valid statement is: 100% of current reported revenue is collaboration/service revenue, no product revenue.

Recursion monetizes its platform in three ways, in order of current importance:

1. Upfront and milestone payments from strategic collaborators. Pharmaceutical companies pay Recursion to apply its platform to targets or disease areas. Payments can include upfront fees, research funding, milestone payments tied to pipeline progress, and royalties on any resulting products.
2. Potential internal pipeline value. Recursion retains some internal programs and could eventually monetize them through out-licensing, partnerships, or commercialization.
3. Data/technology licensing? Recursion does not operate a broad enterprise SaaS licensing model. If anything, it enters into bespoke collaborations. This is a key difference from Schrödinger’s software-like revenue.

Traditional gross margin is not meaningful here because there is no cost of goods sold for a drug product. Recursion has large R&D and G&A expenses that dwarf collaboration revenue.

• FY2023 total revenue: $44.6 million [NPSV]

This figure is from my prior knowledge of Recursion’s FY2023 disclosures and is not independently verified in this session. Primary source should be Recursion 10-K FY2023, EDGAR CIK 0001601830; exact accession/exhibit not verified.

• FY2023 net loss: $328.1 million [NPSV]

Primary source should be Recursion 10-K FY2023; exact accession/exhibit not verified.

• Q1 2024 revenue: $13.5 million [NPSV]

Primary source should be Recursion 10-Q Q1 2024; exact accession/exhibit not verified.

• Q1 2024 net loss: $91.8 million [NPSV]

Primary source should be Recursion 10-Q Q1 2024; exact accession/exhibit not verified.

• Operating margin: negative; the operating loss is substantially larger than revenue. If the above figures are correct, FY2023 net loss is roughly 7.4× revenue, meaning operating leverage is not present. This is a clinical-stage, cash-burning platform company.

Only pharmaceutical collaborators and, to a lesser extent, public research grantors pay Recursion today. The two named large pharma partners from SEC filings are:

• Genentech, a member of the Roche Group — entered a collaboration with Recursion to deploy the Recursion OS across neurology and oncology indications. The agreement was announced in December 2021. Recursion reported receiving $150 million upfront [NPSV] and was eligible for significant milestone payments.

Source: Recursion 8-K/company disclosure; EDGAR 8-K search: sec.gov ; accessed 2026-04-26; exact accession/exhibit not verified in this session.

• Bayer AG — entered an earlier platform collaboration; the deal was smaller than the Genentech/Roche deal and is described in Recursion’s 10-K.

Source: Recursion 10-K FY2023, business and collaboration agreement section; EDGAR URL above; exact accession/exhibit not verified in this session.

Do not infer additional customers from news or marketing pages. From the issuer’s own filings, the only large, named strategic collaborators I can cite are Genentech/Roche and Bayer. Recursion may have other research collaborations, but the material, SEC-disclosed partners are those two.

The customer base is pharmaceutical and biotechnology companies that need drug discovery capacity or target validation. Recursion is also indirectly selling to its own future pipeline via internal programs.

• Large pharma strategic partner: Roche/Genentech — has paid upfront and will pay milestones if programs progress.
• Large pharma partner: Bayer — earlier collaboration, also milestone-based.
• Internal discovery organization: Recursion itself is a “buyer” of its own platform for internal pipeline programs.

Do not infer other named customers. The SEC filings name the material partners; everything else is not source-bound.

Q2 2026 (ended 06-30, 10-Q filed 08-05) revenue $7.7M (−60% YoY from $19.2M — lumpy collaboration milestones, Roche/Genentech-driven). R&D $89.6M (down from $128.6M YoY); net loss $131.0M (improved ~24% YoY). Cash + restricted cash $556.8M; runway guided into early 2028. FY26 cash opex guidance lowered to <$375M (from <$390M). Market cap ~$1.58B at $3.50/share; 534.12M shares; accumulated deficit −$2.324B. Gross margin −98% (lab/compute costs expensed against thin milestone revenue). [10-Q CIK 0001601830]

1. Discovery-cycle compression = the value claim. REC-1245 (RBM39 degrader): biological discovery → development candidate in ~18 months (vs ~4–5-yr industry norm). REC-7735 (PI3Kα H1047R): delivered in ~10 months using only 242 compounds (vs industry norm of thousands, years). Compressing ~3 years of discovery saves ~3 yrs of fully-loaded medicinal-chemistry R&D (≈$25–50M/yr/program → $75–150M/program), but — the honest counterweight — speed does NOT by itself raise Phase 3 success rates, which is where the real value lives.
2. Cost of the moat vs moat it defends. Cumulative accumulated deficit $2.324B (the capital shareholders funded to build the flywheel) vs ~$400M external validation to date (Roche-Genentech ~$213–216M, Sanofi ~$134M, NVIDIA $50M) → the flywheel has consumed ~5.8× what external partners have paid in. The moat is real but expensive; partners have validated it but not yet funded it.
3. R&D intensity / flywheel cost. Q2 R&D $89.6M vs revenue $7.7M = 11.6× R&D/revenue; annualized R&D ~$358M vs FY25 revenue ~$66M = ~5.4×. This is a platform investment posture, not a profitable business.
4. Runway + dilution cost-of-moat. $556.8M cash ÷ ~$94M/qtr guided cash opex = ~6 quarters (early 2028, matching guidance). $300M ATM facility ÷ $1.58B market cap = ~19% potential dilution at current prices.

\[

\$44.6\text{M} / 2 = \$22.3\text{M}

\]

Below are the key parameters for modeling RXRX. Every exact figure marked [NPSV] must be checked against the filing. None are confirmed in this session.