X-Risks Leaderboard — leaderboard
Extreme-risks evaluation leaderboard for frontier models, using 3C3H scoring over biology, chemistry, and cybersecurity risk-domain questions.
Metric: 3C3H Score (%). Source: huggingface.co. Status: saturation imminent. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O1 (2024-12-17) | 29.09 |
| 2 | O3 Mini (2025-01-31) | 27.73 |
| 3 | O1 Mini (2024-09-12) | 20.78 |
| 4 | GPT-4o (2024-08-06) | 18.92 |
| 5 | Llama 3.3 70B Instruct | 17.83 |
| 6 | Qwen 2.5 72B Instruct | 16.6 |
| 7 | GPT-4o Mini (2024-07-18) | 15.51 |
| 8 | Claude 3.5 Sonnet (20241022) | 14.45 |
| 9 | Claude 3 Haiku (20240307) | 13.06 |
| 10 | QwQ 32B-Preview | 11.03 |
Interactive version: theaggregate.ai/benchmark?slug=x-risks-leaderboard · How the rankings work · Data refreshed daily, snapshot 2026-07-22.