What is Gemini for Science?

At Google I/O 2026, Google announced Gemini for Science — a collection of experimental AI tools designed to help researchers generate hypotheses, conduct testing, and understand scientific literature [citation:10].

Three Primary Features

Hypothesis Generation

Digs through millions of scientific papers to generate theories or challenges. Claims are "deeply verified and supported by clickable citations" for absolute rigor [citation:10].

Computational Discovery

An agentic search engine that can generate thousands of tests and experiments much faster than manual methods [citation:10].

Literature Insights

AI-powered chat that searches scientific literature and generates digestible versions through reports, infographics, audio, or video overviews [citation:10].

Science Skills Tool

Also launching as part of Gemini for Science, Science Skills can scour insights from 30+ major life science databases and tools, allowing researchers to "perform complex and often manual workflows in minutes rather than hours" [citation:10].

Availability

Google is gradually opening access to these experimental tools starting today. Researchers can fill out a form on Google Labs website. Enterprise organizations can access via Google Cloud [citation:10].

Pricing

Experimental access — pricing to be determined.

Pros

  • Hypothesis generation with citations
  • Agentic search for experiment design
  • 30+ life science database integration
  • Multiple output formats (reports, infographics, audio, video)
  • Enterprise access via Google Cloud

Cons

  • Experimental — may have limitations
  • Access requires application
  • Pricing not announced
  • Primarily life sciences focused initially

Who Should Use It?

Perfect for: Scientific researchers, life science professionals, academic institutions, and enterprise R&D teams.

Verdict

Gemini for Science represents Google most focused AI application for domain-specific research. The combination of hypothesis generation, computational discovery, and literature insights could accelerate scientific discovery significantly [citation:10].

Rating: 4.4/5 - AI for the scientific method.