OpenAI released mathematical results from its frontier model on October 6, 2026, saying it aims to push the frontier of human knowledge and enable further progress in mathematics. It is the company's first major release in mathematics since its initial research efforts in the field.
OpenAI reported the average result used the equivalent compute of roughly three hours of ChatGPT Pro thinking, measured on the Pro usage benchmark on ChatGPT. That compares with no prior figure provided by the source.
The results are built on internal frontier models and target advancing mathematical understanding through formalizations and transparency. Availability begins with a GitHub repository, initially for the math community and researchers.
"We’re releasing a broad range of new mathematical results produced by an internal frontier model," said OpenAI. "We want this progress to push the frontier of human knowledge and enable further progress in mathematics."
The announcement follows consultations with the Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study. OpenAI said it is committed to improving the quality of the papers via citations, mathematical exposition, and presentation of the results for better understanding.
OpenAI did not say when the model that produced these results will be released, and it raised the open question of how to responsibly disclose major scientific advancements. OpenAI will continue to act on feedback from the community and update its standards for future disclosures of major scientific advancements.
Source: openai
