Anthropic is expanding Claude access for scientists with a new team-plan program and broader AI for Science credits.
The company says it is opening 10,000 seats for scientists around the world for one year. Standard seats will be free. Premium seats with 5x usage limits will cost $15 per month. Anthropic says it intends to expand beyond the first 10,000 seats over time.
The program is tied to Claude Science, Anthropic’s research workbench product, and to its AI for Science program. It moves the company’s science strategy from product availability into subsidized usage for academic and nonprofit research teams.
The team-plan program starts with PIs
Anthropic says principal investigators or equivalent leaders at academic or nonprofit research institutions can apply for the Claude team plan for scientists. Once verified, they can add researchers in their lab to the plan.
That qualification matters. The free and discounted team seats are not a general consumer promotion. They are structured around research groups, institutional verification, and one-year access.
Anthropic says labs that need more usage can also apply to the AI for Science program for up to $50,000 in credits per project. The company says any researcher is eligible to apply for those project credits.
The model-access limits remain important
Anthropic is also keeping dual-use restrictions in place. For now, researchers working in biology and chemistry will still be limited to Opus-class models. Anthropic says Claude Fable models will continue to block professional biology and drug-development queries because of potential dual-use risks.
The company says it is working with the U.S. government on an access program for life-sciences professionals to use Mythos-class models for life-sciences research and development.
That makes the access story more nuanced. Anthropic is expanding use, but not making every model available to every scientific domain. In sensitive areas, the company is treating model capability as something that needs gated deployment.
Science access is becoming a competitive surface
Frontier labs are increasingly treating scientists as a strategic user group. The reasons are direct: scientific workflows create hard reasoning tasks, valuable benchmarks, credible public examples, and demand for long-context, tool-using agents.
For researchers, the practical question is not whether a model is impressive in a demo. It is whether the tool can produce auditable artifacts, work with real analysis environments, respect institutional data constraints, and stay affordable enough for repeated use.
Anthropic’s access program is one answer to the affordability problem. The next test is what researchers publish after using it.
