A competition advancing the physical reasoning capabilities of open weight AI models, built on a benchmark of original, AI-hard physics problems.
- Prizes
- Over $100,000 USD
- Problems
- Olympiad to Research level
- Format
- Teams train and fine-tune open weight models
About the Competition
The AIPhO Prize challenges the open-source community to build AI models capable of advanced physical reasoning. Throughout the competition, teams will train and fine-tune open-weight models to solve highly complex problems. By pushing the boundaries of what these systems can achieve, AIPhO aims to establish a definitive new benchmark for physical intuition in AI.
Thanks to the generous support of our sponsors, the competition features a total prize pool exceeding $100,000 USD.
Who We Are
Our team consists of IPhO medalists, national physics olympiad team coaches, olympiad problem writers, PhD students, and active researchers.
Join Us
We are currently hiring problem writers who are capable of writing original problems in their area of expertise that current state-of-the-art LLMs are not able to solve. Pay is competitive.
Sponsors
We are welcoming additional sponsors to help support the competition — please reach out if your organization is interested in partnering with us.
Contact
For any questions, email [email protected].