HLE

Frontier multi-disciplinary academic knowledge.

Get in touch Read research

2,158Public tasks

57.9%Highest score

The HLE leaderboard

Tests models on 2,158 frontier-level, closed-ended academic questions requiring genuine subject mastery rather than retrieval.

Paper Data Code Sample task

Model

Score

Muse Spark 1.1

Muse Spark 1.1xHigh

57.9%

Opus 5.5

Opus 5.5Max

57.4%

GPT-6 Astra

GPT-6 AstraxHigh

55.0%

Fable 5.1

Fable 5.1Max

54.6%

GPT-6.1 Sol

GPT-6.1 SolMax

54.5%

0%

10%

20%

30%

40%

50%

60%

70%

80%

Pass@1