On June 30, 2026, Anthropic launched Claude Science in beta, describing it as an AI workbench for scientists that consolidates fragmented research tooling into one integrated, agentic environment. Users work with a coordinating agent that has access to more than 60 curated skills and over 100 specialized scientific databases, including UniProt, PDB, and Ensembl, pre-configured for domains such as genomics, proteomics, structural biology, and cheminformatics.
Anthropic emphasizes reproducibility: outputs such as 3D protein structures, genome tracks, and chemical visualizations carry a complete code history so results can be validated and reproduced, and reviewer agents flag and correct errors in citations and calculations. The platform also manages computational scaling from laptops to HPC clusters and on-demand GPUs without moving sensitive datasets.
The beta is available to Claude Pro, Max, Team, and Enterprise subscribers on macOS and Linux, with discounted Team pricing for academic and nonprofit research labs. Alongside the launch, Anthropic opened an AI for Science grant program supporting up to 50 projects with 30,000 dollars in API credits each (plus 2,000 dollars of compute credits from Modal); applications close July 15, 2026, and funded projects run September 1 through December 1, 2026.
The launch signals that frontier labs now see whole professions, not just chat, as product categories: after coding tools, scientific research is the next domain getting a dedicated, agent-native application layer.