Anthropic unveils Claude Science, a unified AI workbench for scientific research
Anthropic has launched Claude Science, an AI-powered research workspace designed to streamline scientific workflows and data analysis.

Anthropic has introduced Claude Science, a new AI-powered workbench designed to help scientists carry out computational research within a single environment, reducing the need to switch between databases, research pipelines and AI tools.
The company clarified that Claude Science is not a new AI model or a specialised biology model. Instead, it runs on the same Claude models already available today, without requiring special access or gated availability.
The launch builds on Anthropic's October 2025 introduction of Claude for Life Sciences and marks the company's broader push to expand beyond AI models into industry-specific workflows. Similar to how Claude Code is positioned as a software development platform, Claude Science aims to serve as a unified workspace for scientific research.
The announcement comes just months after OpenAI introduced GPT-Rosalind, a biology-focused AI model for scientific reasoning. Unlike Claude Science, which is available in beta to Pro, Max, Team and Enterprise subscribers, GPT-Rosalind was released as a research preview limited to qualified U.S. enterprise customers and select partners.
Google DeepMind, meanwhile, is taking a different approach by combining its proprietary scientific models, including AlphaFold and AlphaGenome, with more than 30 life sciences databases through its Gemini for Science platform. Together, the three companies are pursuing distinct strategies for the scientific AI market, with Anthropic offering broader access, OpenAI targeting enterprise customers, and Google relying on its in-house research models.
Anthropic has also launched a research programme supporting up to 50 Claude Science projects with credits of up to $30,000 each. The initiative will initially focus on postdoctoral and graduate research in biomedical sciences, with applications open until July 15, 2026.
How it works
Claude Science connects to more than 60 scientific databases and includes pre-built toolkits for disciplines such as genomics, protein structure analysis and chemistry. The AI assistant acts as a research coordinator, capable of creating specialised sub-assistants to handle different stages of a project.
The platform also includes a separate AI-based verification system that reviews citations and calculations before research is prepared for publication.
Anthropic said Claude Science can also run on a laboratory's own infrastructure, allowing organisations to keep research data on-premises instead of sending it to Anthropic's servers.

Project Lighthouse: Bezos-led Prime Video overhaul puts AI front and centre
AMD doubles down on AI infrastructure with Helios launch to rival Nvidia
Google burns cash for first time as AI spending pushes 2026 capex to $205 billion
Tesla sells more cars but makes less money: Why Musk’s AI pivot is squeezing profits
OpenAI launches Presence to bring AI agents into customer support and enterprise workflows
