Anthropic Launches Claude Science, an AI Workspace for Researchers

Anthropic launched Claude Science on June 30, 2026 — a dedicated AI workspace for researchers that brings together more than 60 scientific databases and computational tools in a single interface. Announced in San Francisco alongside a busy week for the company (the Claude Sonnet 5 release the same day, and Fable 5’s return the day after), it marks Anthropic’s most direct push into scientific research and the pharmaceutical industry so far.

What Is Claude Science?

Anthropic describes Claude Science as a customizable app that integrates the tools and packages researchers most often use, produces auditable artifacts, and provides flexible access to computing resources. In practice, that means a researcher can query dozens of otherwise-incompatible scientific databases, run computational tools, and keep a verifiable record of how each result was produced — all in one workspace instead of stitching sources together by hand.

CEO Dario Amodei framed the ambition simply: a general-purpose technology that helps make sense of biology “in its full complexity, better.” Anthropic also said it plans to pursue its own drug-development programs — moving beyond selling tools into doing discovery work itself.

Why It Matters

• Speed of research: researchers quoted at launch suggested automating the information-gathering stage could increase the pace of experimentation dramatically, with some projecting drug-development timelines could compress from 10–15 years toward 2–5.
• Auditability: producing verifiable, auditable artifacts addresses one of the biggest objections to using AI in regulated research.
• Strategy: together with the biology safeguards built into Claude Fable 5 (and the program that lifts them for select biomedical researchers), it signals a coordinated push into life sciences.

Early Caveats

Launch coverage also flagged real limitations: the beta was reported as Mac/Linux only, some chemical-database integrations were incomplete, and researchers cautioned that hallucinations or missed nuance in regulatory guidance carry real consequences in drug development — so skilled human oversight remains essential rather than optional.

August 2026 Update: Free Claude for 10,000 Scientists

On August 27, 2026, Anthropic expanded the program. It opened 10,000 free and discounted Claude Team seats for scientists: standard seats are free, and premium seats with five times the usage limits cost $15 per month, with pricing fixed for one year. Applicants must be a principal investigator (or equivalent) at an academic or nonprofit research institution and complete a verification form.

Two more changes landed alongside it. Anthropic’s AI for Science credit program, previously limited to the biological sciences, now covers compute-intensive fields including mathematics, physics, and engineering. And in partnership with the U.S. government, Anthropic is establishing an access program that lets vetted life-sciences professionals use its Mythos-class models — the tier behind Claude Fable 5 — for life-sciences research and development, with the first participants already enrolled. The same day, Anthropic previewed the Model Hardware Standard, which lets agents operate lab instruments directly.

The Bigger Picture

Claude Science caps an eventful stretch for Anthropic: Sonnet 5 became the new default model on June 30, Fable 5 was restored on July 1 after a brief export-control suspension, and on July 2 the company published details of Fable’s cyber safeguards along with a proposed industry-wide jailbreak severity framework. For the current model lineup, see Claude Models Explained.

Frequently Asked Questions