GPT-6-Astra takes first on MathArena with 90% expected performance, using just 12k tokens per question
scaling01 · x · 2026-09-06
GPT-6-Astra claims first place on MathArena with a massive 90% expected performance. The model is also notably token-efficient, averaging only 12k tokens per question on BrokenArXiv and ArXivMath.
More from Models
- User demo: GPT-6 Astra builds a professional horse-site UI in minutes from a simple prompt — aziz4ai · 2026-09-06
- Unverified Claims: GPT-6 'Astra' Upgraded Its Own Tooling and Nailed Robot Arm IK Control — repligate · 2026-09-06
- Unverified: GPT-6 Astra Reportedly Completes 279 of 280 Enterprise Agent Jobs — ryanshrout · 2026-09-06
- Anthropic co-founder Tom Brown: frontier-level intelligence is ~20x cheaper in a year — victor_explore · 2026-09-06
- When robot data hits VLM pipelines, VLA vs WAM distinction will vanish — ChongZzZhang · 2026-09-06
- Claim: "3.8" was secretly available a day before its official release — cgarciae88 · 2026-09-06