Benchmarks · reasoning
Humanity's Last Exam
Humanity's Last Exam: frontier expert-level questions across disciplines; accuracy.
146 models · 278 results · data as of 2026-10-29 · artificialanalysis.ai ↗
Leaderboard
Opus 5.5 max
More from this benchmark
3 published measuresScore vs any published measure3 measures: sub-scores, splits, counts
Breakdown
darker is better within each column · click a column to sort| Model | |||
|---|---|---|---|
| Opus 5.5 max, fallback | 950,250 | 1,944,392 | 75,381,566 |
| Gemini 4 Carbon high | 72,667,886 | 2,344,900 | 107,960,277 |
| Fable 5.1 max, fallback | 951,726 | 1,440,360 | 67,639,293 |
| Gemini 4 Argon high | 20,131,633 | 1,654,973 | 32,392,108 |
| Sonnet 5.5 max, fallback | 936,055 | 1,346,542 | 115,825,919 |
| Opus 5 max | 954,049 | 1,815,530 | 56,134,785 |
| GPT-6 Astra max | 680,333 | 594,815 | 15,586,913 |
| GPT-6.1 Sol max | 680,137 | 571,077 | 16,033,363 |
| GPT-5.6 Sol max | 882,090 | 469,832 | 28,312,132 |
| MiMo V2.6 Pro | 708,452 | 5,690,126 | 42,846,712 |
| Muse Spark 1.3 max | 657,368 | 659,017 | 65,228,591 |
| Opus 4.8 max | 963,294 | 2,034,369 | 68,817,988 |
| GPT-6 Sol max | 680,631 | 319,190 | 25,380,172 |
| Gemini 3.7 Flash high | 690,352 | 1,064,276 | 19,960,550 |
| Gemini 3.8 Flash high | 690,352 | 1,308,562 | 51,487,896 |
| Kimi K3 max | 846,369 | 517,205 | 55,939,930 |
| GPT-6.1 Terra max | 680,500 | 398,574 | 40,636,801 |
| Muse Spark 1.1 xhigh | 955,835 | 1,913,408 | 52,370,220 |
| GPT-5.5 xhigh | 667,111 | 507,920 | 26,726,938 |
| Muse Spark 1.2 xhigh | 966,156 | 2,004,209 | 48,804,908 |
| Haiku 5.5 max | 905,424 | 1,449,939 | 187,657,153 |
| Grok 4.6 xhigh | 1,093,928 | 527,524 | 43,989,122 |
| GPT-5.4 xhigh | 666,849 | 656,440 | 35,111,485 |
| DeepSeek V4.1 Flash 1017 max | 706,900 | 623,468 | 81,493,967 |
| Grok 4.7 xhigh | 718,874 | 691,601 | 110,539,347 |
All results
278 rows| # | Model | Effort | Humanity's Last Exam |
|---|---|---|---|
| 1 | Opus 5.5 | max | 61.4% |
| 2 | Gemini 4 Carbon | high | 60.4% |
| 3 | Fable 5.1 | max | 59.1% |
| 4 | Fable 5.1 | xhigh | 58.7% |
| 5 | Opus 5.5 | xhigh | 57.5% |
| 6 | Gemini 4 Argon | high | 57.1% |
| 7 | Fable 5.1 | high | 55.9% |
| 8 | Opus 5.5 | high | 55.6% |
| 9 | Sonnet 5.5 | max | 55.0% |
| 10 | Opus 5 | max | 54.9% |
| 11 | Opus 5.5 | medium | 54.7% |
| 12 | GPT-6 Astra | max | 54.7% |
| 13 | GPT-6 Astra | xhigh | 54.6% |
| 14 | Opus 5 | xhigh | 54.4% |
| 15 | Fable 5.1 | medium | 53.8% |
| 16 | GPT-6 Astra | high | 53.1% |
| 17 | GPT-6.1 Sol | max | 52.9% |
| 18 | Opus 5 | high | 52.8% |
| 19 | GPT-6 Astra | medium | 52.7% |
| 20 | GPT-6.1 Sol | xhigh | 52.6% |
| 21 | GPT-6.1 Sol | high | 51.4% |
| 22 | Opus 5 | medium | 51.3% |
| 23 | Sonnet 5.5 | xhigh | 50.0% |
| 24 | GPT-6.1 Sol | medium | 49.9% |
| 25 | GPT-5.6 Sol | max | 49.5% |
| 26 | MiMo V2.6 Pro | 49.4% | |
| 27 | GPT-6 Astra | low | 49.2% |
| 28 | Fable 5.1 | low | 48.9% |
| 29 | Muse Spark 1.3 | max | 48.7% |
| 30 | Opus 4.8 | max | 48.7% |