Researchers at Stanford University have announced the debut of Biomni, an AI-powered multi-skilled biomedical research agent capable of designing and executing complex research workflows. The system, described in a paper published in Science, can read scientific literature, form hypotheses, select datasets, write code, and interpret results, effectively automating the tedious mechanics of science that often slow discovery.
"If you think of an agent as a carpenter, a carpenter without tools is just a carpenter who can talk," said Jure Leskovec, the Alfred and Rebecca Lin Professor and professor of computer science in the School of Engineering, and senior author of the paper. "With Biomni, we give the carpenter a set of tools so it can build."
Born for impact
Biomni was designed so that a scientist can describe a research problem in simple, natural language. The agent then reads the literature, forms hypotheses, chooses datasets and tools, writes code, interprets results, and suggests next-stage experiments. "Biomni is able to understand a simple question like, 'Why are these patients responding differently to the drug?'" said Kexin Huang, a former doctoral student in Leskovec's lab who recently earned his PhD. "Then it digs in, doing a lot of the scientific legwork."
In one real-world test, a user uploaded more than 450 files of continuous glucose monitoring, food intake, and physical activity data and asked for interesting and plausible hypotheses. Biomni cleaned and unified the data, generated visualizations, and identified patterns linking food intake and body temperature in 40 minutes. Leskovec estimates the same work would have taken a human researcher 60 hours or more.
Biomni also provides full citations and tracks every step of its work. That traceability, the researchers argue, makes the resulting science more rigorous and more reproducible.
Innovation apace
Biomni was trained on full-text papers, code, and data from bioRxiv, a preprint server for biomedical findings. It layers in 150 specialized biomedical tools, 105 software packages, and 59 databases spanning all 25 biomedical subdomains defined by bioRxiv, from genetics to neurology. That integration makes it a practical example of AI Agents & Automation that can carry out complex, multi-step experimental tasks.
Leskovec explained that an inverse relationship exists between the volume of scientific information and the pace of discovery. "The hurdle in biomedical science is not intelligence or ideas; it is mechanics," he said. "It's this laborious stuff that slows innovation. Biomni can do this work in minutes."
Human in the loop
Both Leskovec and Huang stress that Biomni will not replace human researchers. "This is not about machines taking over science, but more about machines becoming a powerful new partner to augment human researchers," Huang said. "With Biomni, scientists have a fast and tireless collaborator that empowers them to focus on the important work of science."
A prototype is already in use by more than 10,000 labs in academia and industry, making it the most widely used AI for Science & Research system in biomedicine. "Biomni is my first research project that has gained wide use by real biologists," Huang said. "To have that impact on how biologists are doing their work has been rewarding. I look forward to seeing where Biomni goes from here."
Why this matters for Science and Research
Biomni shows that specialized AI agents can take over the repetitive, time-consuming mechanics of research-data cleaning, literature review, and code generation-that often consume weeks or months. The glucose monitoring example demonstrates that the same tasks can be compressed into minutes, freeing scientists to focus on experimental design and critical judgment. The system's traceability also addresses reproducibility head-on, a persistent concern in the life sciences. For researchers in biomedicine, adopting a co-scientist like Biomni means getting from a question to a testable hypothesis far faster, without sacrificing the rigor that peer review demands.
Your membership also unlocks:
