Claude Science, Anthropic's new AI research workbench, launched in beta today with more than 60 scientific databases, code, and compute in a single workspace.
Anthropic didn't just ship a smarter lab notebook. On Tuesday it unveiled Claude Science, a workspace that bundles more than 60 scientific databases, code, and compute into a single environment for working scientists, per broader launch coverage, and said it will also use the platform to run its own pre-clinical drug programs aimed at diseases the pharmaceutical industry has long ignored.
The pivot is the more revealing half of the announcement. Anthropic's head of life sciences, Eric Kauderer-Abrams, said the company is "developing its own pre-clinical drug programs focused on neglected diseases." Those are illnesses that disproportionately affect low-income populations and have not been a profitable target for large pharmaceutical companies, and the implied bet is that the world's neediest patients are also the most defensible beachhead for an AI lab trying to build a vertical life-sciences business.
The workbench itself is the more visible piece. Claude Science is a browser-based environment pre-configured for genomics, single-cell, proteomics, and cheminformatics. It can render scientific artifacts such as three-dimensional protein structures, genome browser tracks, and chemistry drawings, and traces every result back to the source code and data environment that produced it, a feature Anthropic is pitching to regulators and reviewers who want to audit how a conclusion was reached. The workbench runs on the same Claude models that have passed Anthropic's responsible-scaling and biosecurity evaluations, the company said, and is available in beta to Claude Pro, Max, Team, and Enterprise customers.
The strategic geometry is harder to miss. Anthropic has been building out its life-sciences operation since October 2025, and the timing of the launch, a workbench plus a drug pipeline announced in the same week, lands as the company is widely described in financial press as IPO-bound. A pre-clinical pipeline aimed at diseases outside the traditional pharma target zone gives Anthropic a narrative moat that doesn't depend on competing for the same high-margin drug targets every well-capitalized AI lab is now chasing. It also gives the company a story to tell public-market investors that is not just "we sell tokens to enterprises."
The neglected-disease framing is not new for Anthropic, but it is the most concrete one yet. CEO Dario Amodei, speaking at Tuesday's launch event, called biology "not as simple as code" but said AI could be "a general-purpose technology for biological complexity," then hedged that "we don't know for sure if that's going to work out." Beta testers reported "significant efficiency gains," per Anthropic's own announcement, though the company did not name the organizations or quantify the gains in the materials reviewed.
Anthropic is also seeding the ecosystem. The company said it will fund up to 50 Claude Science AI for Science projects with as much as $30,000 in credits each, with applications open through July 15 and award notifications by July 31. Whether the workbench becomes the default lab for working biologists, or just another portal in a crowded field, will depend on what the beta cohort actually publishes, and on whether Anthropic's own drug programs ever reach a clinical stage.