insitro sells validated drug targets

insitro is selling R&D output, not a lab copilot, which puts it in the same budget bucket as early drug programs rather than software tools. The customer does not pay for seats or prompts. It pays for targets that look causal in human disease, or molecules that can move into preclinical development. That is why insitro combines patient data, cell models, gene perturbation, and modality design into one workflow aimed at handing a pharma team something it can advance.1 2 3 4 5

• The Gilead deal was structured around discovering up to five NASH targets, with upfront, milestone, and royalty economics. That is classic biotech value capture. The payment scales with whether a target becomes a program and whether a program becomes a drug, not with software usage.2
• The Bristol Myers Squibb collaboration shows the same logic in ALS. The relationship expanded from one nominated target in December 2024 to three targets by March 2026, and insitro is now working across both oligonucleotide and small molecule approaches. The product is a pipeline asset package, not a chat interface.3 4
• This also explains why insitro looks different from infrastructure companies like Benchling or broad AI assistants for scientists. Benchling helps teams manage experiments and data. insitro uses proprietary data and models to decide which gene to hit, then what kind of drug to build against it.1 5

The next step is deeper movement from target nomination into owned and partnered drug candidates. As insitro adds modality design capabilities and advances programs like CTRO-1013, more of the company’s value will come from proving that its biology engine can repeatedly turn human data into medicines that survive the long path to the clinic.4 5 6