Anthropic Launches Claude Science for Autonomous Scientific Research
Anthropic released Claude Science, a flagship product for scientists that autonomously runs experiments — from writing code to identifying drug candidates. Available to all paid subscribers.
On June 30, 2026, Anthropic introduced Claude Science — a full-featured scientific research product elevated to the same status as Claude Code for programming and Claude Cowork for team collaboration. This isn't a plugin or an experiment: Anthropic positions it as a flagship direction.
What Claude Science Can Do
The system works autonomously: a scientist provides a high-level task, and Claude Science writes the code, runs it on powerful compute clusters, and calls on specialized tools for genomics, chemistry, and protein biology. Every step is reproducible — researchers can trace and verify any intermediate decision.
In a demo, the system independently identified new drug candidates for treating phenylketonuria — a rare genetic disorder. Harvard physicist Matthew Schwartz, who tested the underlying Opus 4.5 model, rated it as equivalent to a second-year PhD student on scientific projects.
How This Differs from Previous Attempts
In October 2025, Anthropic released "Claude for Life Sciences" plugins — a more limited offering. Claude Science is a fundamentally different step: a standalone product with its own interface and deep integration into the research workflow. Anthropic also announced it will use Claude Science internally for research into rare and neglected diseases.
The product is available to all paid Claude subscribers as of June 30, 2026. The primary focus is pharma and biotech, though the scope of application isn't limited to those fields.
What This Changes
Until now, AI in science was used primarily as a literature search and summarization tool. Claude Science takes the next step — turning the model into an active participant in the experiment. If this positioning holds up in practice, it could significantly accelerate the early stages of biomedical research — the costliest and most unpredictable phase of drug development.
Source: www.technologyreview.com
Free course
Stop reading about AI — start building with it
The free Claude Code course: your first site, tool or game — no coding. No upsells, no cross-sells — nothing to buy here.
Start free →▌ Related guides

Author
Evgenii Arsentev
PhD · Chief Executive Officer, digital health
Articles · Latest articles