Anthropic launches Claude Science, an AI workbench for scientists

Anthropic has introduced Claude Science, an AI workbench designed for scientific research. The application integrates tools commonly used by researchers—such as Jupyter, R, and PubMed—into a single environment where scientists can analyze literature, execute multi-step research, and refine figures and manuscripts. It is available in beta for Claude Pro, Max, Team, and Enterprise users on macOS and Linux, with support for remote access via SSH or HPC login nodes.

Claude Science features a generalist coordinating agent with access to over 60 curated skills and connectors for fields including genomics, single-cell analysis, proteomics, structural biology, and cheminformatics. The platform generates auditable artifacts, embedding the exact code, environment details, and message history required to reproduce figures and results. It also manages compute resources, allowing jobs to scale from a single GPU to hundreds across a lab's existing infrastructure or on-demand services like Modal.

The system includes a reviewer agent that checks citations, calculations, and figure-code consistency. It connects to scientific databases such as UniProt, PDB, Ensembl, and ChEMBL, and integrates with NVIDIA's BioNeMo Agent Toolkit for access to models including Evo 2, Boltz-2, and OpenFold3. Early users from organizations including Manifold Bio, the Allen Institute, and UCSF have reported using Claude Science for tasks such as CRISPR screen design, protein structure prediction, and epidemiological analysis.

Anthropic is also offering up to $30,000 in credits for up to 50 Claude Science AI for Science projects, with applications open through July 15, 2026. A discounted Team plan is available for academic and nonprofit research labs.