Google launches Gemini for Science with AI tools to streamline scientific research workflows

Google Introduces Gemini for Science to Transform Research Workflows

Google announced Gemini for Science during its I/O 2026 conference, unveiling a collection of AI-powered experimental tools designed to streamline scientific research workflows and tackle a growing challenge in modern science 1

1

2

2

. The initiative addresses what Google describes as a paradox facing researchers today: collective knowledge is expanding so rapidly that individual scientists struggle to see the full picture, with manual data analysis often taking weeks or even months 2

2

. These AI tools aim to serve as a force multiplier for scientific work, allowing researchers to focus on identifying and tackling the most impactful problems.

Three Core Features Drive the New Era of Discovery

Gemini for Science includes three primary features that map directly to the scientific method. Hypothesis Generation helps researchers during the initial steps by digging through millions of scientific papers to develop theories or identify challenges 1

1

. Google emphasizes that claims generated by this tool are "deeply verified and supported by clickable citations" for absolute rigor 1

1

. Once a hypothesis is established, Computational Discovery acts as "an agentic search engine" that can generate thousands of tests and experiments much faster than manual methods 1

1

. The third feature, Literature Insights, functions as an AI-powered chat that scours scientific literature and produces more digestible versions through reports, infographics, or even audio and video overviews 1

1

.

Science Skills Integrates Over 30 Life Science Databases

Alongside the core experimental tools, Google is launching Science Skills, a specialized bundle that integrates insights from over 30 major life science databases and tools including UniProt, AlphaFold Database, AlphaGenome API, and InterPro 2

2

. Using these skills on agentic platforms like Google Antigravity allows researchers to perform complex workflows in structural bioinformatics and genomic analyses in minutes rather than hours 2

2

. In early testing, Google's research teams used Science Skills to perform a complex analysis that normally takes hours in just minutes, leading to novel insights about potential mechanisms for a rare genetic disease caused by mutations in the AK2 gene 2

2

.

Enterprise Organizations Gain Access Through Google Cloud

Google is gradually opening access to these experimental tools starting today through Google Labs, where interested researchers can register 1

1

2

2

. Beyond individual researchers, Google is offering enterprise-grade solutions for scientific and industrial R&D through Google Cloud 2

2

. Enterprise organizations including BASF, Klarna, Daiichi Sankyo, Bayer Crop Science, and the U.S. National Labs are already using these tools in private preview 2

2

. BASF is using AlphaEvolve to optimize supply chains, while Klarna leverages it to enhance machine learning models 2

2

. Several validation papers based on these tools have been published, with ERA and Co-Scientist research papers appearing in Nature 2

2

. Google plans to expand access to more organizations in the coming months as it continues demonstrating the value of these tools to accelerate scientific discovery.