Anthropic STEM Fellowship 2027 (up to $167,000)
Overview
The Anthropic STEM Fellowship 2027 brings researchers and experts from science and engineering into Anthropic for a duration of four months. Participants work in small pods alongside peers from related fields, combining their domain expertise with frontier models and internal tooling. Fellows tackle grand challenges by identifying open problems in their disciplines, building datasets and simulations to test and train Claude, and discovering where models fall short.
Who can apply
Applicants must hold a PhD in a STEM field and possess several years of experience, such as a post-doc or faculty role. Candidates need hands-on experience using Claude and Claude Code in their daily workflows and must be excited about the potential for artificial intelligence to reason through complex questions in their fields. Additionally, applicants should be comfortable operating in a fast-paced environment to deliver concrete results and must hold existing authorization to work in the United States for the duration of the program. Prior machine learning experience is helpful but not required, while scientific judgment and a willingness to learn quickly are prioritized.
What you get
Participants receive between $100,000 and $167,000 USD for the four-month fellowship, which is equivalent to an annualized rate of $300,000 to $500,000 USD, along with other benefits.
Requirements
Fellows are responsible for setting grand challenges by identifying long-horizon open problems where Claude falls short of expert judgment and defining what progress entails. Responsibilities also include building comprehensive datasets, simulations, and rigorous evaluations to test whether models can plan experiments, interpret data, and reason about mechanisms. Furthermore, participants apply Claude to open problems using test-time compute strategies and integrate the scientific tools and software relied upon by their specific field.
Key dates
Applications remain open until October 28, 2026, with offers planned for November 20, 2026. The selection process typically includes a take-home exercise and discussions with the research team regarding research interests and potential project directions.