CAIS releases HLE-Diamond, a refined 1,000-question Humanity's Last Exam; GPT-6 Astra reportedly scores 82.9% with tools
Around Sept 22, 2026 the Center for AI Safety released HLE-Diamond, a cleaned 1,000-question subset of Humanity's Last Exam (500 reasoning, 500 knowledge). Secondary coverage reports that GPT-6 Astra scored 82.9% with web and code tools, the first HLE-series result above 75–80%. The prediction-market contract on a ≥75% score by end-2026 jumped from 14% to 87%.
Key facts
- HLE-Diamond: 1,000 questions (500 reasoning + 500 knowledge), a refined subset of HLE
- Reported: GPT-6 Astra 82.9% with tools (web + code), per Octagon/Kalshi coverage; not verified on the CAIS page, whose charts are not in its text
- Previous best on the main HLE reported at about 65% (Claude Fable 5.1)
- Kalshi 'HLE ≥75% by end of 2026' market moved from 14% to 87%
- HLE-Rolling was updated on Sept 17, 2026
What happened
CAIS produced a higher-quality core of HLE after criticism that some original questions had wrong or ambiguous answers. The first reported frontier score is far above the main-HLE scores from earlier in 2026.
Why it matters
If confirmed, the benchmark once billed as 'the last exam' is close to saturation within about 20 months of release. Confidence is medium until the score is confirmed on the CAIS leaderboard.
Changelog
- 2026-09-29: created
Related events
Sources (2)
id: 2026-09-22-hle-diamond · updated 2026-09-29 · open in the interactive timeline