Bristol Myers Squibb is bringing Chai Discovery’s AI into its antibody programs. The two companies unveiled a strategic collaboration on Thursday that puts the startup’s models at the center of antibody research.
Under the deal, BMS will weave Chai’s molecular folding and de novo design models into its drug discovery portfolio. Chai frames the result as a continuously learning system that designs candidates with AI rather than manual screening. Its platform behaves like a computer-aided design suite for biology, predicting and reprogramming molecular interactions and generating antibody sequences from scratch.
Finding a viable antibody the old way can take one to two years of repeated lab screening. Chai says its tools can compress that early phase from months into weeks, especially for flexible membrane proteins and epitopes that have resisted drugging.
The deal follows Chai’s $400M Series C announced last month, which valued the company at $3.8B. It also expands a pharma roster that already includes Eli Lilly, Pfizer, and Novartis, underlining how foundation-model-driven discovery stacks are becoming standard industry infrastructure. Founded in 2024 by alumni of OpenAI, Meta FAIR, Stripe, and Absci, Chai counts Joshua Meier as its CEO.
Financial terms were not disclosed. The partnership lands as Reuters reports pharma broadly doubling down on AI to cut costs and timelines.
