benchmarks / hle
Humanity's Last Exam
accuracy. Status active. Contamination risk low. Maintainer Scale / Epoch collection. Source.
| Model | Score | Date | Kind |
|---|---|---|---|
| GPT-6 Astra | 0.5479999999999999 | 2026-09-03 | observation |
| Claude Fable 5.1 | 0.465 | 2026-09-01 | observation |
| Gemini 3.1 Pro Preview | 0.4644 | 2026-02-19 | observation |
| Gemini 3.8 Flash | 0.44520000000000004 | 2026-09-02 | observation |
| GPT-5.4 Pro | 0.4432 | 2026-03-05 | observation |
| Muse Spark | 0.4056 | 2026-04-08 | observation |
| Gemini 3 Pro Preview | 0.37520000000000003 | 2025-11-18 | observation |
| GPT-5.4 | 0.3624 | 2026-03-05 | observation |
| Claude Opus 4.7 | 0.362 | 2026-04-16 | observation |
| Claude Opus 4.6 | 0.3444 | 2026-02-05 | observation |
| GPT-5 Pro | 0.3164 | 2025-10-07 | observation |
| GPT-5.2 | 0.278 | 2025-12-11 | observation |
| GPT-5 | 0.2532 | 2025-08-07 | observation |
| GPT-5 | 0.2532 | 2025-08-07 | observation |
| Claude Opus 4.5 | 0.252 | 2025-11-24 | observation |
| Kimi K2 5 | 0.24370000000000003 | 2026-01-27 | observation |
| GPT-5.1 | 0.2368 | 2025-11-13 | observation |
| Gemini 2.5 Pro Preview 06.05 | 0.2164 | 2025-06-05 | observation |
| O3 | 0.20320000000000002 | 2025-04-16 | observation |
| GPT-5 Mini | 0.19440000000000002 | 2025-08-07 | observation |
| O3 | 0.192 | 2025-04-16 | observation |
| Claude Opus 4.6 | 0.19 | 2026-02-05 | observation |
| Gemini 2.5 Pro Exp 03.25 | 0.1816 | 2025-03-25 | observation |
| O4 Mini | 0.1808 | 2025-04-16 | observation |
| Gemini 2.5 Pro Preview 05.06 | 0.17800000000000002 | 2025-05-06 | observation |
| O4 Mini | 0.1428 | 2025-04-16 | observation |
| Claude Sonnet 4.5 | 0.13720000000000002 | 2025-09-29 | observation |
| Gemini 2.5 Flash Preview 04.17 | 0.1208 | 2025-04-17 | observation |
| Claude Opus 4.1 | 0.1152 | 2025-08-05 | observation |
| Gemini 2.5 Flash Preview 05.20 | 0.10960000000000002 | 2025-05-20 | observation |