Insilico puts Liquid AI's 2.6B model into MMAI Gym, setting SOTA in retrosynthesis

helloiamleonie · x · 2026-09-03

Insilico Medicine placed Liquid AI's 2.6B-parameter language model into its MMAI Gym training environment, where it now sets the state of the art in single-step retrosynthesis.

The result is a notable AI-for-science case: a small general-purpose LM, combined with a chemistry-specific training setup, beating existing methods on chemical synthesis planning.

Original post →

More from Research

Research channel →