For Centuries, the Scientific Method Has Driven Human Progress
At Google, our mission is deeply rooted in building acceleration tools. We believe that a new era of discovery will come not from narrow, specialized models, but from general agents that empower researchers in all scientific fields.
That’s why we’re introducing Gemini for Science, a collection of scientific tools and experiments designed to expand the scope and precision of scientific research.
A Force Multiplier for Human Ingenuity
Today, science faces a paradox: our collective knowledge is growing so quickly that it is becoming increasingly difficult for individual scientists to see the big picture. Scientific breakthroughs often rely on making creative connections between data, but the time required to do this manually can take weeks or even months. AI can help eliminate this bottleneck and serve as a force multiplier for scientific work by tackling complex tasks. This allows researchers to focus on identifying and addressing the most influential scientific problems and directions that would drive progress.
Gemini for Science’s experimental tools at Google Labs include three primary prototypes designed to tackle such tasks.
Hypothesis Creation
Created with a co-scientist, ideation is the heartbeat of science, but no human can summarize the millions of articles published annually. Hypothesis Generation fills this gap by simulating the scientific method: it works with researchers to define a research challenge and then uses a “tournament of ideas” with multiple agents to generate, discuss, and evaluate hypotheses. To ensure complete accuracy, claims are thoroughly verified and supported with clickable quotes.
Computational Discovery
Built with AlphaEvolve and ERA (Empirical Research Assistance), scientific progress is often limited by the number of hypotheses we can realistically test with computational experiments. Computational Discovery, an agent discovery engine, is a prototype that solves this problem by generating and evaluating thousands of code variants in parallel. This allows scientists to test novel modeling approaches – for complex areas such as solar forecasting or epidemiology – that would take months to navigate manually.
Literature Insights
Created with Google NotebookLM, understanding scientific literature is a central part of all research trips. Literature Insights searches scientific literature and organizes the results into tables with custom, searchable attributes for parallel analysis. Researchers can use chat to discover nuances embedded in their curated corpus and create highly accurate artifacts such as reports, slide decks, infographics, and audio and video overviews. With the power of NotebookLM, Literature Insights helps synthesize insights from diverse papers, identify research gaps, and uncover areas of opportunity.
Starting today, we will gradually release access to these experiments. Visit labs.google/science to register your interest.
Beyond individual experiments, we also make these advanced AI capabilities available to enterprise organizations via Google Cloud. Our corporate solutions for scientific and industrial research and development are already being used in private preview by a number of partners to increase real-world impact. Companies like BASF use AlphaEvolve to optimize their supply chains and Klarna uses it to improve their machine learning models. In parallel, organizations such as Daiichi Sankyo, Bayer Crop Science and the US National Labs (as part of the US Department of Energy’s Genesis Mission) are using Co-Scientist to accelerate their research and address fundamental scientific challenges. These enterprise-class tools show significant value in their current preview phase. We’re excited about the breakthroughs our partners are making 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 research by ERA and co-scientists is published today in Nature.
A Scientific Workbench on Your Desktop
As part of Gemini for Science, we are also launching Science Skills, a dedicated package that integrates insights from over 30 major life science databases and tools, including UniProt, AlphaFold Database, AlphaGenome API and InterPro. Using these capabilities on agent platforms like Google Antigravity enables researchers to complete complex and often manual workflows like structural bioinformatics and genomic analysis in minutes rather than hours.
Our research teams using Science Skills have already seen this acceleration in practice. In early testing, our team used scientific skills to perform complex analysis that typically takes hours into minutes. This led to new insights into possible mechanisms of a rare genetic disease caused by mutations in the AK2 gene.
For more information, click here.
“`
