OpenAI's Astra Model Solves 10 Math Problems, Costing ~$2,000 Each
imjustnewatai · x · 2026-08-01
OpenAI announced ten major advances in mathematics and theoretical computer science using its next major model, Astra. The breakthroughs tackle open problems in areas like sphere packing, coding theory, and lattice cryptography that had seen no progress for at least a decade.
An internal version of Astra generated the mathematical arguments, collaborating with humans to prepare 249-page manuscripts, and then formalized each argument in Lean. OpenAI estimates the solution-search tokens across all ten problems cost roughly $2,000 at current API rates.
If these papers survive community review, it suggests scientific discovery is becoming a scaling problem. OpenAI also plans to provide frontier models to 100,000 researchers through 2027.
Related event: OpenAI's Internal Model Astra Cracks 10 Major Math Problems(99 posts)→
More from Models
- Leaking deep residual vectors into early layers may fix state tracking in frozen LLMs, zero retraining — burny_tech · 2026-09-18
- Models know they're reward hacking in 50-96% of rollouts, Goodfire's activation monitors catch it in real time — burny_tech · 2026-09-18
- Qwen3.8-Omni-Flash cuts overlapping-speech error rate from 88% to 3% and drops audio API pricing 98% — karminski3 · 2026-09-18
- Qwen3.8-Omni-Flash: meeting ASR errors cut from 88% to 3%, API prices down 98% — karminski3 · 2026-09-18
- Gemini 3.8 live beats gpt-live-1 on some benchmarks, say insiders — bosmeny · 2026-09-18
- 105 planted bugs benchmark: Unbiased's Pareto scores 30.7 for just $4.81 — PawelHuryn · 2026-09-18