Chai Discovery, a startup building AI models that design antibodies and other biological molecules, has closed a $400 million Series C at a $3.8 billion valuation, tripling its price tag just seven months after raising a $1.3 billion Series B in December 2025. The round, led by Index Ventures, marks one of the fastest valuation climbs in the AI-for-biology sector this year and underscores how quickly investors are pricing in the promise of models that can propose and score novel proteins the way large language models generate text.
A Round That Moved Fast
New investors in the July 14, 2026 raise include Bain Capital Ventures, Battery Ventures, and Baillie Gifford, joining existing backers OpenAI, Thrive Capital, Sequoia Capital, and Kleiner Perkins. The pace of the fundraising — a tripled valuation in under eight months — mirrors the boom-time dynamics seen in large language model startups, except applied to a company whose core product is not a chatbot but a molecular design engine meant to shorten the years-long, multi-billion-dollar process of discovering new drugs.
What Chai Discovery Actually Builds
Founded by former Google DeepMind and Meta AI researchers, Chai Discovery trains foundation models on protein structure and sequence data to predict how candidate antibodies and small molecules will fold, bind, and behave before a single one is synthesized in a lab. The company already has paid deployments with Eli Lilly, Pfizer, and Novartis, according to people familiar with the contracts, giving it a foothold inside three of the largest drug makers in the world at a stage when many AI-biology startups are still working with academic labs or mid-size biotechs.
Why Big Pharma Is Buying In
Traditional antibody discovery can take a research team 18 months or longer to screen thousands of candidates in wet-lab assays before finding one that is both effective and safe enough to advance into animal studies. Chai Discovery’s pitch to pharmaceutical partners is that its models can narrow that search space computationally, flagging a small number of highly promising candidates before any physical synthesis begins. Executives at partner companies have described the tools as a way to compress early discovery timelines rather than replace the clinical trial process itself, which remains governed by the same regulatory requirements as any other drug candidate.
The Skeptical Case
Not every biotech investor is convinced computational protein design has matured to the point that justifies an eight-month valuation triple. Critics of the AI-biology funding wave note that few of these platforms have yet produced a drug that has cleared Phase 3 trials, and that predicting binding affinity in silico still frequently diverges from how a molecule behaves in a living organism. Some venture investors who passed on the round have said valuations across the sector are increasingly being set by comparison to other AI startups rather than by traditional biotech milestones like clinical data or regulatory filings, a dynamic that could leave later investors exposed if wet-lab results lag the hype.
OpenAI’s Growing Bet on Biology
OpenAI’s continued participation as a repeat investor is notable: the AI lab has increasingly backed startups applying its research and compute-adjacent expertise to life sciences, treating biology as one of the highest-value domains for frontier models beyond text and code. For Chai Discovery, that relationship extends beyond capital, with the startup reportedly drawing on techniques adapted from large-scale foundation model training to build its protein models.
What Comes Next
With fresh capital, Chai Discovery is expected to expand its roster of pharmaceutical partnerships and invest further in wet-lab validation capacity to test its computational predictions against real biological outcomes, a step the company says is necessary to build the track record that skeptics are demanding. Whether the startup’s models actually shorten the path from digital prediction to an approved drug will likely take years to answer, since even a promising computationally designed candidate must still clear the same preclinical safety studies and multi-phase human trials required of any other therapeutic. For now, the round cements Chai Discovery as one of the best-capitalized players in a crowded field of AI-drug-design startups racing to prove that molecular foundation models can do for biology what large language models did for text.
