Anthropic launches Claude Science workbench for researchers

Anthropic launched Claude Science on Tuesday, introducing an AI workbench—a software environment for computational research—designed to centralize computational research. The tool connects to more than 60 scientific databases, sparing researchers from moving between databases, pipelines, and tools. Anthropic clarified that Claude Science is “not a new AI model and not a more capable model for biology. It runs the same Claude models already available to everyone today (including Claude Opus 4.8), with no special access and no gating.”

The launch represents a strategic pivot for Anthropic, which is increasingly focusing on vertical, workflow-level products to own the operating layer for specific industries. The company is betting its growth on these workflow-level tools rather than raw model capability. To support adoption, Anthropic will support up to 50 Claude Science projects, providing up to $30,000 in credits. Applications for this support are open through July 15, 2026, with award notifications sent by July 31. The projects will run from September 1 to December 1, 2026, with an early focus on fields across biomedical research.

The workbench features automated project management and a separate fact-checker AI. This verification step is designed to address risks associated with AI-assisted writing, which can lead to fabricated citations and unverifiable stats. For reproducibility, the platform generates scientific figures alongside the code used to create them. Early users of the tool include the Allen Institute and the UCSF Brain Tumor Center. At UCSF, researchers used the workbench to accelerate germline analysis—the genetic analysis of inherited variants—of glioma, achieving independently validated results in less time than previously required.

The launch intensifies competition in the scientific AI sector, placing Anthropic against OpenAI and Google DeepMind. The three companies are pursuing distinct distribution strategies:

• Anthropic: Offering broad subscription access to Claude Science for Pro, Max, Team, and Enterprise users.
• OpenAI: Taking an enterprise-gated approach with GPT-Rosalind, a specialized model for biological reasoning released in April that remains limited to qualified enterprise customers in the U.S.
• Google DeepMind: Leveraging its foundational science models within the Gemini for Science platform, which integrates more than 30 life science databases.

Why it matters

Anthropic is increasingly betting its growth on vertical, workflow-level products rather than just raw model capability, aiming to own the operating layer for specific industries.