MIRAGE — leaderboard
Metric: Violent completion rate, $C_1$ (%) (self-reported). Source: benchmarklist.com. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3 | 26 |
| 2 | Llama 3.3 70B Instruct | 24 |
| 3 | Qwen 2.5 72B Instruct | 22 |
| 4 | Mistral Large | 19 |
| 5 | GPT-4o | 12 |
| 6 | Claude Opus 4 | 8 |
Interactive version: theaggregate.ai/benchmark?slug=mirage · How the rankings work · Data refreshed daily, snapshot 2026-07-22.