For centuries, the scientific method has been the greatest driver of human progress. Google’s mission is deeply rooted in building the tools that accelerate it. We believe that a new era of discovery will not come from narrow, specialized models, but from general agents that empower researchers in all scientific fields.
That’s why we introduce Gemini for Science, a collection of scientific tools and experiments designed to expand the scale and precision of scientific exploration.
Power amplifier of human ingenuity
Today, science faces a contradiction. Our collective knowledge is growing rapidly, making it increasingly difficult for individual scientists to see the big picture. Scientific advances often depend on making creative connections between data, which can take weeks or even months to do manually. AI can help eliminate this bottleneck and double the power of scientific research by handling complex tasks. This allows researchers to focus on identifying and addressing the most impactful scientific questions and directions that will drive progress.
Google Labs’ Gemini for Science experimentation tools include three main prototypes designed to handle such tasks.
Hypothesis generation with co-scientists: Ideas are the heart of science, but humans can’t synthesize the millions of papers published each year. Hypothesis generation fills this gap by simulating the scientific method. Collaborate with researchers to define research questions and use multi-agent “idea tournaments” to generate, discuss, and evaluate hypotheses. To ensure absolute rigor, claims are deeply verified and supported by clickable quotes. Computational Discovery Built with AlphaEvolve and ERA (Empirical Research Assistance): Scientific progress is often limited by the number of hypotheses that can realistically be tested with computational experiments. Computational Discovery, an agent research engine, is a prototype that solves this problem by generating and scoring thousands of code variations in parallel. This allows scientists to test new modeling approaches in complex fields such as solar forecasting and epidemiology that would take months to navigate manually. Literature Insights Built with Google NotebookLM: Understanding the scientific literature is at the core of all research efforts. Literature Insights searches scientific literature and structures the results into tables with custom searchable attributes for side-by-side analysis. Researchers can use chat to uncover nuances based on a curated corpus and create high-fidelity deliverables such as reports, slide decks, infographics, audio and video summaries, and more. Powered by NotebookLM, Literature Insights helps you integrate findings across papers, identify research gaps, and uncover areas of opportunity.
Starting today, we will gradually open up access to these experiments. Visit labs.google/science to register your interest.
In addition to conducting individual experiments, we also offer these advanced AI capabilities to enterprise organizations through Google Cloud. The enterprise-grade solution for scientific and industrial research and development is already being used by various partners in private preview to drive real-world impact. Companies like BASF are using AlphaEvolve to optimize their supply chains, and Klarna is leveraging it to power its machine learning models. In parallel, organizations such as Daiichi Sankyo, Bayer Crop Science, and the U.S. National Laboratory (as part of the U.S. Department of Energy’s Genesis Mission) are using Co-Scientist to accelerate research and address fundamental scientific challenges. These enterprise-grade tools are demonstrating great value in their current preview stage. We are excited about the breakthroughs our partners are unlocking and look forward to expanding access to more organizations in the coming months.
Several validation papers have already been published based on these and other tools. The ERA and Co-Scientist research paper was published today in Nature.
science workbench on desktop
As part of Gemini for Science, we are also launching Science Skills, a specialized bundle that brings together insights from more than 30 leading life science databases and tools, including UniProt, AlphaFold Database, AlphaGenome API, and InterPro. Using these skills with agent platforms such as Google Antigravity, researchers can perform complex and often manual workflows such as structural bioinformatics and genomic analysis in minutes instead of hours.
Our research team using Science Skills has already seen this speedup in the wild. In early tests, our team used our science skills to perform complex analyzes that would normally take hours in minutes. This provides new insights into the potential mechanisms of rare genetic diseases caused by mutations in the AK2 gene.
