Google has launched Science Skills in Google Antigravity, a bundle of AI-powered research tools designed to accelerate scientific discovery across biology, genomics, and computational modeling. The new capabilities integrate insights from more than 30 life science databases, allowing researchers to complete workflows that typically take hours in just minutes.

What Exactly Is Google Antigravity's Science Bundle?

Google Antigravity is the company's agent-first development platform, and it now includes a dedicated Science Skills bundle as part of the broader Gemini for Science initiative announced at Google I/O 2026. The Science Skills bundle integrates data from major life science resources including UniProt, AlphaFold Database, AlphaGenome API, and InterPro, giving researchers access to a unified interface for complex biological queries.

The Science Skills bundle works alongside three experimental research tools available through Google Labs. These tools are designed to support the full research lifecycle, from initial literature review through hypothesis testing to computational discovery.

How to Use Google's New Science Research Tools

• Literature Insights: Searches scientific literature, structures results into data tables, and generates reports, slide decks, infographics, and audio and video overviews to help researchers synthesize findings across papers and identify research gaps.
• Hypothesis Generation: Uses a multi-agent "idea tournament" system to help researchers define challenges, generate research ideas, and evaluate hypotheses with clickable citations and potential flaw identification.
• Computational Discovery: Generates and scores code variations in parallel using AlphaEvolve and Empirical Research Assistance (ERA), designed to test modeling approaches in areas like solar forecasting and epidemiology.

How Much Faster Are These Tools, Really?

Google's research teams tested the Science Skills bundle in early trials and completed an analysis that normally takes hours in just minutes, leading to insights about possible mechanisms for a rare genetic disease caused by mutations in the AK2 gene. This dramatic speed improvement could reshape how researchers approach complex biological problems, potentially accelerating the pace of scientific breakthroughs across multiple disciplines.

The speed gains matter because they free researchers to focus on interpretation and hypothesis refinement rather than waiting for computational results. For academic labs with limited resources, this efficiency could level the playing field against well-funded institutions.

Who's Already Using These Tools?

Google is collaborating with more than 100 academic and research institutions to validate the new systems, including Stanford University School of Medicine, Imperial College London, and The Francis Crick Institute. The company's trusted tester community includes PhD students, industry researchers, and Nobel laureates, signaling serious scientific credibility behind the initiative.

Enterprise-grade versions are already in private preview with major organizations. BASF is using AlphaEvolve to optimize supply chains, Klarna is leveraging it for machine learning models, and Daiichi Sankyo, Bayer Crop Science, and the U.S. National Labs through the Department of Energy's Genesis Mission are using Co-Scientist.

"The next era of scientific discovery will be aided by AI agents acting as force multipliers for human ingenuity," stated Pushmeet Kohli, Vice President at Google DeepMind and Chief Scientist at Google Cloud.

Pushmeet Kohli, Vice President at Google DeepMind and Chief Scientist at Google Cloud

When Can Researchers Access These Tools?

Access to Gemini for Science is opening gradually through Google Labs starting in May 2026, with users able to register interest. The rollout is intentionally phased to allow Google to gather feedback and refine the tools based on real-world research workflows. Researchers can expect to gain access over the coming months, though the exact timeline depends on demand and testing progress.

Google has also created pilots with major conferences including ICML, STOC, and NeurIPS for agentic peer review and scientific validation tools, including Paper Assistant Tool and ScholarPeer. These partnerships suggest the company is thinking beyond individual researchers to reshape how the scientific community validates and publishes findings.

What Does This Mean for the Future of Research?

The integration of AI agents into scientific workflows represents a fundamental shift in how research gets conducted. By automating literature synthesis, hypothesis generation, and computational testing, these tools could democratize access to advanced research capabilities. Smaller labs and institutions in developing countries could potentially compete with well-resourced research centers, accelerating global scientific progress.

James Manyika, Senior Vice President at Google-Alphabet, emphasized the strategic importance of this direction, noting that agentic science and tools to accelerate scientific progress represent a major area of focus for the company. This suggests Google sees scientific AI as a long-term competitive advantage, not just a near-term product feature.

The research papers on ERA and Co-Scientist have already been published in Nature, providing peer-reviewed validation of the underlying technology. This academic credibility is crucial for adoption among researchers who need confidence that AI-assisted tools produce reliable results.