MIRAGE: leaderboard
Five years after the discovery of persistent anti-Muslim bias in large language models, most evaluations remain confined to single-turn prompt completion, a setting that no longer reflects how frontier LLMs are deployed. We introduce MIRAGE (Muslim-Identity Reasoning and Agentic Generation Evaluation), a benchmark of 1,200 prompts spanning three deployment-realistic conditions: direct
Metric: Violent completion rate, $C_1$ (%) (self-reported). Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4 | 8 | #163 |
| 2 | GPT-4o | 12 | #343 |
| 3 | Mistral Large | 19 | #710 |
| 4 | Qwen 2.5 72B | 22 | #499 |
| 5 | Llama 3.3 70B | 24 | #579 |
| 6 | DeepSeek V3 | 26 | #321 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mirage · How It Works · Data refreshed daily, snapshot 2026-10-20.