@rohanpaul_ai: $2K of tokens costs less than sending 2 people to a conference, and that gap is what changes the calculation for open p…

X AI KOLs Following Models

Summary

OpenAI reports that an internal version of its next major model (Astra) solved 10 long-standing open problems in math and theoretical computer science for roughly $2,000 in tokens, with formal Lean certificates.

$2K of tokens costs less than sending 2 people to a conference, and that gap is what changes the calculation for open problems. An internal version of Astra, OpenAI's next major model, resolved 10 long-standing open problems in math and theoretical computer science for about $2K in tokens. complete with formal Lean certificates.
Original Article
View Cached Full Text

Cached at: 08/03/26, 09:52 PM

$2K of tokens costs less than sending 2 people to a conference, and that gap is what changes the calculation for open problems.

An internal version of Astra, OpenAI’s next major model, resolved 10 long-standing open problems in math and theoretical computer science for about $2K in tokens.

complete with formal Lean certificates.

OpenAI (@OpenAI): An internal version of our next major model produced 10 new results on long-standing open problems in mathematics and theoretical computer science, using roughly $2,000 worth of tokens at GPT-5.6 Sol API rates.

Similar Articles

Ten advances in mathematics and theoretical computer science

Simon Willison's Blog

OpenAI used an internal model, Astra, to solve ten mathematical problems that had stalled for over a decade, spending under $2,000 per problem and releasing Lean 4 formalizations and a paper. The results prompt reflections on AI's role in mathematics.