Loading data...
AA Humanity's Last Exam: leaderboard
Artificial Analysis independent evaluation of HLE: 2,500 expert-vetted frontier questions across math, sciences, and humanities.
Metric: Accuracy (%). Source: artificialanalysis.ai. Status: years away from saturation. 602 models tracked.
Top models
| # | Model | Score |
|---|
| 1 | Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 59.13 |
| 2 | Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback) | 58.71 |
| 3 | Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback) | 55.93 |
| 4 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | 55.47 |
| 5 | Claude Opus 5 (Adaptive Reasoning, Max Effort) | 54.87 |
| 6 | GPT-6 (Max) | 54.68 |
| 7 | GPT-6 (xHigh) | 54.59 |
| 8 | Claude Opus 5 (Adaptive Reasoning, Xhigh Effort) | 54.4 |
| 9 | Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback) | 53.8 |
| 10 | GPT-6 (High) | 53.06 |
| 11 | Claude Opus 5 (Adaptive Reasoning, High Effort) | 52.83 |
| 12 | GPT-6 (Medium) | 52.73 |
| 13 | Claude Opus 5 (Adaptive Reasoning, Medium Effort) | 51.3 |
| 14 | GPT-5.6 Sol (Max) | 49.49 |
| 15 | GPT-6 (Low) | 49.21 |
Interactive version: theaggregate.ai/benchmark?slug=aa-humanity-s-last-exam · How It Works · Data refreshed daily, snapshot 2026-09-05.