Variational AI, Merck team up on drug discovery; deal worth up to $349 million

Variational AI, a Vancouver-based generative AI drug discovery company founded by machine learning researchers from MIT, Caltech, Google Research, Microsoft Research, and D-Wave Quantum, has built its Enki™ platform on a foundation model trained on internal data, a curated version of all publicly available data, and proprietary generative models. This collaboration marks a notable partnership with a major pharmaceutical company, Merck, which is known as MSD outside the United States and Canada. Under the agreement, Variational AI will fine-tune its platform on Merck's proprietary data to work on two targets chosen by Merck, while Merck secures exclusive rights to develop and commercialize any resulting compounds. The deal structure — an upfront payment plus milestones totaling up to $349 million — suggests a pay-for-performance approach that aligns incentives, though the actual payout will depend on hitting undisclosed milestones. The collaboration reflects Merck's stated interest in using AI to improve efficiency, speed, and quality of candidates earlier in the discovery process, as noted by Robert M. Garbaccio of Merck Research Laboratories. For Variational AI, the partnership provides both validation and capital, and its CEO Handol Kim framed it as a framework that could redefine the unit economics of drug discovery. The two undisclosed targets mean the specific therapeutic areas remain confidential, but the deal size indicates meaningful commitment from Merck.