← Back to the wire

GPT-6 Astra gives mathematicians a breather, and OpenAI says that's by design

AchievementBenchmarkSep 10, 2026

OpenAI's GPT-6 Astra ranked first on ulam.ai's ErdosBench, scoring 3.23 and solving 106 of 226 open math problems, 43 completely. Benchmark developer Przemek Chojecki described it as a 5%-10% gain over Sol, which solved 78 problems. OpenAI chief scientist Jakub Pachocki said the company deliberately did not prioritize math, focusing instead on recursive self-improvement and automated alignment research. Compared to Sol, Astra showed stronger scientific writing and fewer overblown claims.

Receipt № 18391 source · awaiting confirmation ◐

Evidence

1source· awaiting independent confirmation

No score is assigned. Sources and their independence are shown in the citation chain below.

Citation chain · 1 source

OpenAICompanySolModelGPT-6 AstraModelJakub PachockiPersonulam.aiCompanyPrzemek ChojeckiPerson
Canonical: https://the-decoder.com/gpt-6-astra-gives-mathematicians-a-breather-and-openai-says-thats-by-design/