Nvidia has launched the BioNeMo Agent Toolkit, a collection of domain-specific tools designed to enable AI agents to perform life sciences research tasks including protein structure prediction, molecular docking, genomic analysis and drug discovery.

The toolkit incorporates Nvidia's BioNeMo, NIM microservices, Parabricks, NeMo and Nemotron technologies. It is available through Nvidia's developer resources page and GitHub.

More than 50 companies are said to be using the toolkit. Anthropic and OpenAI are integrating it into their platforms, while Dassault Systèmes, Databricks, Lilly, Schrödinger and Snowflake are among those adopting it for scientific workflows.

The University of Washington's Institute for Protein Design collaborated with Nvidia on the toolkit, with the partnership producing a 2x performance improvement for the RosettaFold3 biodesign model compared to its prior generation.

"Frontier models are the brains. BioNeMo is the scientific toolbox. Together, they give AI agents the skills of a PhD research assistant and the speed of a supercomputer," said Jensen Huang, founder and CEO of Nvidia.

The toolkit is designed to allow general-purpose AI agents to execute scientific computing tasks such as virtual screening, genomic analysis, protein binder design, biomedical research and medical imaging analysis. Nvidia states the toolkit can compress some screening timelines from days to minutes.

Lab automation companies including Thermo Fisher, Tecan and Automata, along with AI cloud providers Baseten, Modal and Nebius, are also part of the ecosystem building on the toolkit.

Nvidia (NASDAQ: NVDA) said the announcement was made at the BIO conference in San Diego.