Merck taps Variational AI to design drug candidates in deal worth up to $349 million
The collaboration covers two undisclosed targets. Its headline value includes possible milestone payments, while the upfront payment remains undisclosed.
Listen to this story
The audio brief
Story brief
3 key pointsThe collaboration gives Merck exclusive development and commercialization rights to compounds designed by Variational AI for two Merck-selected targets. Variational AI will use Enki, adapted with Merck’s proprietary data, to design and optimize small-molecule candidates; the agreement’s value could reach $349 million, but that figure includes future milestones and is not the payment at signing. Merck frames the work as an effort to improve early discovery, with no results yet reported and the targets undisclosed.
- 01
Variational AI will receive an upfront payment, but the amount and milestone breakdown were not disclosed.
- 02
Merck’s Robert M. Garbaccio cited efficiency, speed and candidate quality as goals for the early-discovery work.
- 03
Variational AI CEO Handol Kim said the collaboration could affect drug-discovery economics; this remains a stated potential, not a demonstrated outcome.
Merck will get exclusive development and commercialization rights to compounds from a new collaboration with Variational AI, which will adapt its generative AI platform using Merck’s private data. Announced November 2, 2026, the agreement assigns the startup two drug-discovery targets and carries a potential value of up to $349 million, including milestone payments.
The assignment is to design and optimize small-molecule candidates—potential drug compounds—against targets selected by Merck.
Merck’s data shapes the molecule-design work
Variational AI will use a fine-tuned version of Enki, its drug-discovery platform, trained on Merck’s proprietary data. Fine-tuning means adapting the model with that additional data for the partner’s assignment.
Variational AI describes the underlying Enki platform as combining a foundation model trained on its internal data and a curated collection of publicly available data with proprietary generative models. The company says the platform creates and optimizes small-molecule leads around a partner’s target product profile: the desired characteristics of a prospective medicine.
For Merck, the stated objective is better work early in discovery. Robert M. Garbaccio, vice president and head of discovery chemistry at Merck Research Laboratories, said the company wants AI to improve the efficiency, speed and quality of candidates at that stage. He described the collaboration’s targets as challenging, without identifying them.
“This is a compelling framework that has the potential to significantly accelerate and redefine the unit economics of drug discovery.”
Handol Kim, CEO of Variational AI
A potential payout, not a payment at signing
The financial terms distinguish money Variational AI will receive upfront from payments it may earn later. The company will receive an upfront payment and is eligible for milestones, bringing the total potential deal value to $349 million. The announcement does not disclose the upfront amount or break down the milestones. The headline figure is therefore not a payment being made now.
The announcement presents faster discovery and better economics as objectives, not demonstrated results from this partnership.
Sources
- biospace.comVariational AI Enters Collaboration with Merck to Apply Generative AI to Drug Discovery
Reader comments
Newest comments first. Replies stay oldest first.