Google · 2026-05-20 · major
Gemini for Science — Google's I/O 2026 Bundle Plugs Co-Scientist, AlphaEvolve, NotebookLM, and 30+ Life-Science Databases Into One Research Workspace
Google bundled Co-Scientist hypothesis generation, AlphaEvolve and ERA computational discovery, NotebookLM literature insights, 30+ life-science databases via Science Skills, and a peer-review assistant into a single labs.google/science workspace for researchers.
Google bundled Co-Scientist, AlphaEvolve, NotebookLM, AlphaFold/AlphaGenome plug-ins, and a peer-review assistant into one Gemini-for-Science workspace.
Key specs
| Life science databases | 30+ |
|---|---|
| Research partners | 100+ |
| Alpha fold users | 3M+ |
| Life science acceleration | hours to minutes |
| Components | 5 |
What is it?
Gemini for Science is a Google I/O 2026 product announcement (May 20, 2026) that packages five experimental research tools into a single labs.google/science workspace. It is aimed at academic researchers, life-science teams and enterprise R&D groups who want one place to hypothesise, search literature, run computational experiments and review papers with Gemini.
How does it work?
The bundle contains: (1) Hypothesis Generation, the multi-agent Co-Scientist that runs an 'idea tournament' and surfaces ranked hypotheses with verified citations; (2) Computational Discovery, an AlphaEvolve- and ERA-powered engine that generates and scores thousands of code variations against an optimisation metric; (3) Literature Insights, a NotebookLM mode that structures papers into tables, reports, slides and audio briefings; (4) Science Skills, native plug-ins into 30+ life-science databases including UniProt, AlphaFold and AlphaGenome; and (5) Paper Assistant Tool, an experimental peer-review reviewer. Early users at Stanford, Imperial College London, The Crick Institute, BASF and Klarna report collapsing multi-hour analyses to minutes.
Why does it matter?
Until now Co-Scientist, AlphaEvolve, NotebookLM and AlphaFold were separate Google products each with their own access loops; bundling them into one workspace turns Gemini into a frontier-lab-grade research stack accessible from a single sign-in. The peer-review assistant is also notable as arXiv just announced a one-year LLM-output ban — Google is moving the opposite direction, embedding LLM review inside the research workflow itself.
Who is it for?
Academic researchers, bioinformaticians, R&D scientists, Google AI Ultra subscribers
Try it
https://labs.google/science