Summary: On August 27, 2026, Anthropic announced 10,000 free or discounted yearlong Claude seats for scientists, with premium seats priced at $15 per month. Eligibility starts with a verified principal investigator at an academic or nonprofit research institution, and AI for Science can provide up to $50,000 in credits per project. Biology and chemistry researchers still face Opus-class restrictions, while access to Mythos through a government program has no stated timeline.
What Changed: A Bigger Door for Scientists
On August 27, 2026, Anthropic laid out a sizable widening of the backing it gives the scientific community. Scientists around the world get access to 10,000 seats for Claude subscriptions, either at no cost or at a discount, for a full year. The vehicle is a new offering called the Claude team plan for scientists. Standard seats carry no charge; premium seats, which raise usage limits by 5x, run $15 per month. Nor is this a short-lived promotion: the company’s stated intent is to stretch the program far past that first 10,000 seats over the coming months.
The announcement stands on earlier groundwork. Back in June 2026, Anthropic released Claude Science, a product that pulls in tools researchers already rely on, turns out auditable artifacts, and gives flexible access to computing resources. Alongside it, the company has been running the AI for Science program, which hands free credits to researchers tackling high-impact work. Together with the new subscription plan and the widened credits offering, they form part of a larger push to “radically accelerate scientific discovery,” in the company’s own words.
How It Works: Verification, Tiers, and Credits
A seat in the Claude team plan for scientists is reserved for a principal investigator (PI) or equivalent at an academic or nonprofit research institution. It begins with a verification form on Anthropic’s site. Verification clears the way to bring the researchers in your lab onto the plan.
Two tiers make up the plan:
- Standard seats: no charge, with standard usage limits.
- Premium seats: $15 per month, with 5x the usage limits of standard seats.
Once your lab burns through its allotted credits and needs more capacity than even premium plans give, the AI for Science program takes applications for as much as $50,000 in credits per project. Eligibility for those credits extends to any researcher, whatever the field.
Anthropic also says the AI for Science program is widening its scope. In the past it mostly backed scientists applying Claude to biological sciences. It now reaches researchers in other scientific fields as well, among them those pursuing ambitious, compute-heavy work—the sort behind progress on the Riemann zeta function and Claude’s work on protein design.
Practical Use Cases: From Math to Protein Design
Two concrete research examples carry the announcement:
- Riemann zeta function: researchers advanced this famous unsolved problem in number theory, in all likelihood deploying Claude for demanding mathematical reasoning and computation.
- Protein design: Claude has gone to work accelerating protein design, a field carrying enormous implications for medicine and biotechnology.
What these examples signal is that Anthropic’s target is not limited to traditional life sciences labs. Mathematicians, physicists, and other fundamental science teams are being courted too. For product builders the message is that AI tools are turning into essentials across scientific disciplines, biology and chemistry no longer excepted.
If your product is an AI tool for scientists, think about how it can carry compute-heavy workflows, plug into the research tools people already use, and emit auditable outputs—the very traits Claude Science stresses. The subscription design also hints that researchers favor pricing they can predict and usage tiers, rather than pay-as-you-go credits on their own.
Limitations and Trade-offs: Model Restrictions and Access Control
One caveat matters here: people doing biology and chemistry research remain confined to Opus-class models. The newer Claude Fable models keep blocking professional biology and drug development queries, on the grounds of potential dual-use risk. A subscription therefore doesn’t put the most advanced models in life scientists’ hands for certain sensitive tasks.
Anthropic states it is working alongside the US government to set up an access program letting life sciences professionals apply Mythos-class models to life sciences research and development. Enrollment has begun for the first participants, though the announcement gives neither a timeline nor details on how it would scale.
To product builders this is a reminder that access controls and safety guardrails belong to product design. Anthropic is walking the line between openness and risk management, and doing it with regulators at the table. Anyone building AI tools for sensitive domains faces kindred trade-offs: handing users real power while keeping misuse out.
Takeaway: A Playbook for Building a High-Value User Community
There’s more than philanthropy in Anthropic’s move. It’s a strategic play to place Claude in scientists’ hands—people positioned to supply high-quality feedback and pull adoption into research settings. Free and discounted seats lower the barrier to entry, credits serve the heaviest users, and out of that comes a community of power users with a stake in the product doing well.
Key lessons for product teams:
- Tiered access: draw users in with a free or discounted entry, then upsell premium tiers carrying higher limits.
- Credits for heavy users: keep a route open for users who blow past even premium limits, so they don’t churn.
- Verification and eligibility: set clear criteria (PI status, for instance) to make sure the right audience shows up.
- Safety guardrails: be upfront about model restrictions, and involve regulators where needed.
What Anthropic’s approach demonstrates is that a deep, engaged community need not come at the expense of commercial sustainability. Design eligibility, usage tiers, and safety measures with care, and the program can serve your users and your product’s long-term growth at the same time.