Biomni is a general-purpose biomedical AI agent developed by Stanford’s SNAP group and collaborators, released as an open-source project on GitHub with about 4,000 stars. It is designed to autonomously execute a wide range of research tasks across biology and medicine, from genomics analysis to clinical data questions.
Biomni combines a large language model with a curated environment of biomedical tools, datasets and software packages. Given a research goal in natural language, the agent plans the work, writes and runs code, queries domain databases, and produces analyses and reports. Its action space covers areas such as genomics, transcriptomics, proteomics, microbiology, pharmacology and clinical informatics, and it can add new tools and data sources at runtime. The project also introduced a benchmark of realistic biomedical tasks used to measure how well agents handle expert-level lab work. Biomni is actively developed, with updates continuing into late 2026.
Biomni is free and open source under the Apache 2.0 license. You run it yourself and only cover the cost of the LLM API calls it makes on your behalf.
Biomni is for biomedical researchers, bioinformaticians and labs that want an AI agent able to carry out multi-step analyses across the standard tools and databases of modern life-science research.