BlueBench - Safety — leaderboard
Metric: Score (%). Source: huggingface.co. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Mistral Large | 88.04 |
| 2 | GPT-4o | 87.84 |
| 3 | GPT-4.1 | 87.49 |
| 4 | GPT-4.1 Nano | 86.67 |
| 5 | GPT-4.1 Mini | 85.96 |
| 6 | Mistral Medium 3 | 84.68 |
| 7 | Llama 3.3 70B Instruct | 83.8 |
| 8 | O4 Mini | 83.55 |
| 9 | Llama 3.2 3B Instruct | 83.37 |
| 10 | Llama 3.2 1B Instruct | 83 |
| 11 | O1 | 82.96 |
| 12 | Pixtral-12B | 82.45 |
| 13 | O3 Mini | 80.61 |
Interactive version: theaggregate.ai/benchmark?slug=bluebench-safety · How the rankings work · Data refreshed daily, snapshot 2026-07-22.