Overview
Valence Discovery uses an AI-driven generative chemistry engine to design novel small-molecule drug candidates. Its few-shot learning approach builds predictive models from limited data to tackle data-scarce targets and optimize multiple properties early in the design process, including potency, selectivity, safety, bioavailability, and synthesizability. It collaborates with academia, CROs, and pharmaceutical companies, with partners such as Charles River Laboratories, Repare Therapeutics, and Servier, typically via upfront payments and milestone-based rewards. In 2023, Recursion acquired Valence for about $47.5 million and positioned it as its AI/ML R&D center in Montreal to continue developing foundational drug-discovery models and leverage Recursion’s datasets and wet-lab facilities.
Work here?Claim your company
Simplify's take
What believers are saying
- Recursion's Q3 2026 filing showed $85.5 million net proceeds from share sales.
- Sanofi advanced its first collaboration candidate toward Phase 1 in 2027.
- Genentech exercised the first neuroscience option in Q3 2026.
What critics are saying
- Valence has vanished as a standalone business, now dependent on Recursion's priorities.
- The 5.9 million exchangeable-share overhang keeps acquisition economics exposed through 2026.
- If Recursion cuts R&D spend or misses milestones, Montreal headcount gets slashed next.
What makes Valence Discovery unique
- Recursion's Valence Labs pairs Montreal AI talent with proprietary chemistry and phenomics datasets.
- Few-shot generative chemistry targets sparse-data programs where competitors need large assay libraries.
- Valence Labs publishes open-source tools like datamol and graphium, recruiting technical credibility.
Help us improve and share your feedback! Did you find this helpful?