Humanity's Sixth Sense: leaderboard
Metric: Pass@1 (%), all rubric criteria; three attempts/task. Source: scale.com. Saturation forecast: Around 2029. 25 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-6 Astra (Max) | 53.6 |
| 2 | GPT-6.1 Sol (Max) | 46.6 |
| 3 | Claude Opus 5.5 (xHigh) | 44.6 |
| 4 | Gemini 3.8 Flash (High) | 41.6 |
| 5 | Claude Fable 5.1 | 40.8 |
| 6 | Gemini 3.7 Flash (High) | 39.8 |
| 7 | Muse Spark 1.3 (Max) | 37.4 |
| 8 | Claude Fable 5 (Max) | 34.5 |
| 9 | Qwen 3.8 Max (Max) | 33 |
| 10 | Gemini 3.5 Flash (High) | 32.6 |
| 11 | Gemini 3.6 Flash (High) | 31.9 |
| 12 | GPT-6 Sol (Max) | 31.2 |
| 13 | Qwen 3.8 Flash (Max) | 30.9 |
| 14 | Claude Opus 5 (Max) | 30.5 |
| 15 | GPT-5.6 Sol (Max) | 30 |
Interactive version: theaggregate.ai/benchmark?slug=humanity-s-sixth-sense · How It Works · Data refreshed daily, snapshot 2026-10-09.