Anthropic unveiled Claude Science on Tuesday, a dedicated product for scientific research and drug development that the company is positioning alongside its existing tools Claude Code and Claude Cowork. The launch signals a direct challenge to Google DeepMind, which has led the field of AI-powered science for the past decade.

What You Need to Know

Claude Science is a standalone product designed to autonomously carry out research tasks, from writing code to running simulations on computer clusters. It builds on an earlier effort called Claude for Life Sciences but now holds equal status with Anthropic's coding and general-purpose agents. The tool is marketed primarily to pharmaceutical companies and academic labs, and Anthropic is also using it internally to pursue its own drug research for neglected diseases.

How Claude Science Differs From Earlier Tools

Anthropic first released plugins for scientific software in October under the heading Claude for Life Sciences. Claude Science, however, is a full-featured agent rather than a set of add-ons. Like Claude Code, Claude Science can take high-level instructions and work through complex tasks independently. It integrates with tools used in genetics, chemistry and protein biology, making it particularly suited for computational biology and drug discovery.

Eric Kauderer-Abrams, Anthropic's head of life sciences, said the product is designed to help scientists run code on powerful computing clusters and maintain reproducibility so every result can be traced back to its source. Claude Science is now available to all paid Claude subscribers.

  • Autonomous research: Works from concise instructions to execute multi-step scientific projects.
  • Cluster integration: Helps manage powerful computer resources needed for large-scale calculations.
  • Traceable outputs: Prioritizes reproducibility so figures and results can be verified.

Anthropic's Scientific Ambitions

The company's decision to elevate Claude Science to the same rank as its coding and coworker agents shows how seriously it takes AI for science. For Abram's team, this is not just a product launch. Anthropic announced it will use Claude Science to pursue its own drug research for rare, neglected diseases, effectively stress-testing the tool while advancing its mission. CEO Dario Amodei, a PhD scientist himself, has long emphasized life sciences as the greatest opportunity for AI to serve humanity's long-term well-being.

For the past decade, Google DeepMind has dominated this space, with its AlphaFold model earning a Nobel Prize for Demis Hassabis and John Jumper. But the fast-moving AI frontier has left DeepMind struggling to keep pace in areas like coding, where Anthropic has surged ahead. Earlier this month, Jumper announced he is leaving DeepMind for Anthropic, a move that strengthens the company's scientific credibility.

Why This Matters

Claude Science enters a market with billions of dollars at stake. Pharmaceutical companies have deep pockets and a growing need for AI tools that can accelerate drug discovery. For Anthropic, landing major pharma contracts could be as important as any scientific breakthrough, especially with an IPO on the horizon and the company approaching profitability. The product also raises the bar for DeepMind, which must now defend its scientific lead or risk being overtaken by a rival that already leads in coding and now targets drug development directly. Meanwhile, scientists gain a powerful new partner capable of functioning at the level of a second-year graduate student, according to research by Harvard physicist Matthew Schwartz published on Anthropic's website. But the tool is not without limitations: it focuses heavily on molecular and cellular biology and requires careful oversight to ensure accuracy and validity.