OpenAI previews 'Astra' — agent teams that solved ten open math problems for about $2,000
OpenAI unveiled a new model family, tentatively called Astra, built to let multiple AI agents work together on hard problems for hours or days. An internal version solved ten open problems in mathematics and theoretical computer science that had resisted experts for at least a decade — including a proof that non-sofic groups exist — with proofs formalized in Lean so machines can check them. Total compute cost at API rates: roughly $2,000. Sam Altman demoed the system to Washington policymakers; it may ship as GPT-6, GPT-5.7 or a separate class, and is expected to be the first model to go through the planned US government pre-release review. (Correction: OpenAI has since walked back its framing that these problems had seen no progress for at least a decade, and named mathematicians say several of the proofs build on published work that was not credited.)
Related stories
Machine learning read the shape of sick brain cells and picked out nine already-approved drugs that calmed them down2026-08-21DeepSeek's cheap workhorse can now see — and on agent tasks that need eyes it says it is close to Anthropic's best
2026-08-21Given four hours and a GPU to improve the way AI is trained, the best agent scored 0.25 out of 1 — and most never tried
2026-08-21ChatGPT can now read your iMessages and send them — and the setting that lets it skip asking is the one OpenAI warns about
2026-08-21Show the robot once — three to twelve seconds — and it gets the job right 59 times out of 100 with no training at all
2026-08-21