Anthropic Launches Claude Science to Automate Biotech and Drug Research

Tools & Releases

Anthropic Launches Claude Science to Automate Biotech and Drug Research

Anthropic has unveiled Claude Science, a standalone flagship AI agent designed to autonomously execute computational biology and drug discovery workflows.

AZAli Zayed · Founder & EditorJuly 1, 20262 min read✓ Independently fact-checked
The quick version
  • Anthropic has launched Claude Science, a standalone flagship product designed to autonomously run computational biology and drug development tasks.
  • The tool is immediately available to all paid Claude subscribers and will also be used internally by Anthropic to research treatments for rare diseases.
  • The release follows the high-profile defection of Nobel laureate John Jumper from Google DeepMind to Anthropic earlier in June 2026.

Anthropic has officially launched Claude Science, a standalone flagship product designed to autonomously execute scientific research workflows, particularly in computational biology and drug development. Announced on June 30, 2026, at an industry event for pharmaceutical and biotechnology executives, the new tool is immediately available to all paid Claude subscribers. According to Anthropic, the system can autonomously perform complex research tasks based on high-level, concise prompts, marking a major step forward in AI-driven scientific discovery.

The launch represents a significant escalation in Anthropic’s positioning against rivals like Google DeepMind, which has historically dominated the AI-for-science sector. DeepMind CEO Demis Hassabis and researcher John Jumper previously won a Nobel Prize for their work on AlphaFold, but Jumper recently left DeepMind to join Anthropic. Unlike Google’s research-centric approach, Anthropic is framing Claude Science as a commercial developer-style tool. It sits alongside Claude Code and Claude Cowork as a core pillar of the company’s product lineup. While scientists are already heavy users of general programming assistants—many of which are tracked in our guide to the best AI coding tools—Claude Science is built specifically to execute and run code within specialized scientific databases and software environments.

Why it matters

According to Eric Kauderer-Abrams, Anthropic’s head of life sciences, Claude Science is not meant to replace coding tools but to build on top of them by helping researchers run simulations and manage data pipelines. The underlying models, including Anthropic’s Opus series, have demonstrated highly advanced reasoning capabilities. Harvard physicist Matthew Schwartz recently estimated that Anthropic’s Opus 4.5 model performs scientific project execution at a level comparable to a second-year graduate student.

What it means for the industry

Anthropic is also putting the tool to work internally, announcing plans to use Claude Science to pursue its own research into treatments for rare and neglected diseases. This move signals a shift in how AI labs operate, moving from pure model providers to active participants in scientific and pharmaceutical discovery. With a PhD scientist at the helm in CEO Dario Amodei, Anthropic appears uniquely positioned to capitalize on the intersection of LLM agents and hard sciences, especially as competitors face challenges translating their research models into highly lucrative commercial developer tools.

Frequently asked questions

Who can access Claude Science?

Claude Science is currently available to all paid Claude subscribers.

What is the difference between Claude Science and earlier Anthropic tools?

While Anthropic previously released scientific database plug-ins under ‘Claude for Life Sciences’ in October, Claude Science is a fully featured, standalone flagship product that can autonomously execute research tasks.

What scientific fields is Claude Science designed for?

The tool is specifically optimized with features and database integrations for computational biology, drug development, and pharmaceutical research.

Our tested pick

If you are looking to automate your development workflows alongside your research, check out our tested list of the best AI coding tools.

Best AI Coding Tools (2026): 7 Tested & Ranked →

Source: MIT Tech Review. Published July 1, 2026.

AZ
Ali Zayed
Founder & Editor · AI Tools Worth

Ali has hands-on tested 50+ AI tools and tracks model releases daily. Every verdict here comes from real, paid usage — never vendor demos or sponsored placements.

AI Tools Worth is independent and unsponsored. Some linked guides contain affiliate links — they never change our verdicts.