Chai Discovery raises $400 million as AI-driven molecular design enters Pfizer and Novartis research programs
Chai Discovery raises $400 million as AI-driven molecular design enters Pfizer and Novartis research programs
On 14 July, Chai Discovery announced a $400 million Series C funding round at a valuation of $3,8 billion. The funding will support computing, data, research, and product development. The announcement followed separate agreements with Pfizer and Novartis.
The search for a new protein drug begins with a basic question: which molecule should researchers synthesize and test? An antibody is a protein that must recognize the intended biological target and bind to it strongly. There are too many possible amino acid sequences to test them blindly in the laboratory.
Chai develops models that predict molecular interactions and propose protein candidates with specified properties. These models help researchers select candidates for initial laboratory experiments. The candidates then undergo target-binding tests, animal studies, and studies in humans.
In June, Biohub models had already produced binding proteins that reached laboratory testing. Chai offers pharmaceutical companies the same early stage of molecular discovery as a practical tool.
On 5 June, Pfizer entered into a licensing agreement with Chai. The company will receive early access to Chai-3 and a separate model that uses Pfizer’s proprietary data and is adapted to its drug discovery process. Pfizer is integrating molecular AI with its proprietary data and drug development workflows.
On 13 July, one day before the funding announcement, Novartis announced a collaboration with Chai to discover therapeutic antibodies against several targets. The companies had already worked together on technical development for more than a year. Novartis will now gain access to Chai-3 for its own therapeutic programs.
Chai’s new funding round will support computing, data, research, and product development. The company plans to expand its computing capacity and the datasets used to build new versions of the model.
The first test of these agreements will be whether they produce candidates that pass laboratory testing and enter clinical programs.