Previewing the Model Hardware Standard

Anthropic has launched a research preview of the Model Hardware Standard (MHS), a shared specification enabling AI agents to operate physical laboratory and manufacturing devices—including microscopes, liquid handlers, robotic arms, and quantum computing equipment—through a standardized driver architecture. Developed initially with HHMI Janelia Research Campus, MHS reduces device integration time from weeks or months to hours or minutes by replacing bespoke hardware integrations with a universal set of primitives (read/write commands) and auto-generated device reference files in natural language. The standard is model-agnostic, compatible with any programmable interface, and accessible via MCP, CLI, or API; early partners in biotech, robotics, and quantum computing report faster iteration cycles and real-time fault detection. A proof-of-concept at Genentech demonstrated MHS coordinating a liquid handler, robotic arm, and plate reader to automate a BCA protein assay, with the broader goal of enabling autonomous, round-the-clock experimental workflows. Anthropic is sharing MHS with select partners to co-develop safety evaluations and best practices before open-sourcing the standard.