Chai Discovery Inc., a company that develops artificial intelligence models to predict interactions between biochemical molecules, today announced it has raised $400 million in new funding, nearly tripling its valuation to $3.8 billion.

Index Ventures led the Series C round alongside Kleiner Perkins, Sequoia Capital and Dimension. New investors joining the investment included Bain Capital Ventures, Battery Ventures, Baillie Gifford, BDT & MSD, Sapphire Ventures, Avra Capital and other notable investors. Existing investors included Thrive Capital, OpenAI, Oak HC/FT, Menlo Ventures and General Catalyst, among others.

The round raises the company’s total funding to about $630 million to date, including $130 million at a $1.3 billion valuation in December.

Chai Discovery builds frontier AI models designed to accelerate drug discovery. They act by predicting and reprogramming molecular interactions, a necessary first step toward fully understanding how chemistry works for pharmaceutical and life sciences companies.

“Tomorrow’s medicines should be designed with the precision, speed and scale of modern engineering, and this support helps us move faster towards that future,” said co-founder and Chief Executive Joshua Meier. “AI drug discovery has moved from promise to deployment.”

The company says that its latest model Chai-3 is a “step change” over the previous generation Chai-2. It reportedly doubles success rates for molecular interaction targets, implying about 35% to 40% hit rates. The company targets enhanced bonding affinity and broader antibody design foundations.

Chai Discovery targets antibodies because they are part of the body’s primary defense against foreign invaders. Antibodies act as custom locks designed to identify and lock down attackers. The problem: There are roughly a quintillion (1 with 18 zeroes) customizations that can uniquely target foreign bodies. The old way of discovering which ones worked involved screening millions of potential molecules and testing them one by one.

To tackle this problem, Chai created an AI system that can perform rapid simulations based on disease targets to design a molecule that fits correctly. It doesn’t need to guess, no pile of random molecules. Instead it curates the vast set of possibilities to generate a likely high-quality candidate.

Recent commercial traction for the company includes a landmark licensing agreement with Pfizer Inc., granting access to Chai-3, as well as an AI model trained on the pharmaceutical company’s proprietary data. It has also signed a customer agreement with the American multinational pharmaceutical company Eli Lilly and Co. and formed a formal collaboration with the Swiss multinational drug company Novartis AG.

Chai faces a very real climb within the drug discovery industry. According to an executive report from AI life sciences firm Excelra Knowledge Solutions Pvt. Ltd., despite about $20 billion being poured into generative AI drug discovery, no AI-discovered drug has been approved yet. According to OncoDaily, a news site about cancer research and treatments, there are more than 173 AI-originated drug programs now in clinical development, up from almost two dozen in 2023. About 15 to 20 are expected to reach working trials in 2026.

The reason is that although AI-driven discovery achieves fairly high accuracy, with Phase I pass rates at 80% to 90%, it drops to about 40% in Phase II. This matches up with the rates of traditional methods.

Ideally, drug discovery and life sciences will need to overcome this hurdle by gaining a better understanding of candidates and their effects as they advance into clinical trials.

Image: Pixabay