Google Antigravity Adds Chemistry Skills That Cut Drug Discovery Work from Days to an Hour

Google's DeepMind team has launched a set of "Cheminformatics Skills" for Google Antigravity. It helps researchers speed up literature search, molecule analysis, and new molecule design in drug discovery.

In a lung cancer drug demo, the AI worked on two targeted drugs, crizotinib and lorlatinib. It sorted through 568 compounds, found a link between molecular rigidification and better drug effects, and designed several improved macrocyclic compounds with synthesis suggestions. Such work used to take days.

The tool combines the reasoning power of large language models with chemistry tools like RDKit, and connects to public databases such as ChEMBL and PubChem. Google stressed the AI only offers testable scientific hypotheses, not finished drugs. The feature is now open, and Google will give up to 10,000 scientists six months of free top-tier usage.