Isomorphic Captures Design Budgets

This comes down to where pharma believes the scarce value sits, in the model that proposes molecules or in the broader machine that runs experiments. Isomorphic is built to win the first budget. Its engine starts from a target, predicts structures and binding sites, generates new compounds, and ranks a short list for synthesis and assays. That is a direct fit for pharma teams that already own wet labs and development infrastructure, but want better candidate design at the front of the funnel.1 2 3 4

• Isomorphic is selling concentrated design capability, not a general lab operating system. It works through bespoke programs with Lilly, Novartis, J&J, and internal teams, and gets paid through upfront fees, research funding, milestones, and royalties. That makes each win large, even without owning robotics or being self serve software.1 2 4 5
• Lila is asking buyers to adopt a much bigger stack. It combines hypothesis generation, experiment planning, robotic execution, and data feedback inside AI Science Factories. That can create a stronger long term data flywheel, but it also requires a customer to trust an outside platform with more of the research workflow and more physical execution.3 6
• The closest proof that pharma will pay for narrower tools is the rest of the market. Schrödinger is deploying Bunsen inside Bristol Myers Squibb on top of familiar workflows, and Recursion after Exscientia offers a fuller therapeutics stack with more than 60 petabytes of data and clinical programs. The budget line is already splitting between design specialists and full stack platforms.7 8 9

The next step is likely a clearer sorting of buyers. Pharma companies with strong internal labs will keep leaning toward the best molecule design layer, while companies that want outsourced discovery capacity will pay for integrated platforms like Lila. If Isomorphic can turn model performance into clinical candidates, it can keep taking premium budgets without ever matching Lila's physical lab footprint.1 3 4 8