IslamicMMLU - Hadith Source Identification: leaderboard
Metric: Accuracy (%) on IslamicMMLU's 1,000 Hadith track questions on naming which of the six canonical collections a hadith (chain of narrators removed) comes from; four-option multiple choice in Arabic, zero-shot, temperature 0, answer letter only (10 new tokens), via each model's API; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 26 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 96.4 | #93 |
| 2 | Claude Opus 4.5 | 94.7 | #79 |
| 3 | Gemini 3 Pro | 93.1 | #77 |
| 4 | Gemini 2.5 Pro | 91.7 | #145 |
| 5 | Claude 3.7 Sonnet | 90.9 | #241 |
| 6 | Claude Sonnet 4.5 | 89.8 | #138 |
| 7 | GPT-5.2 | 89.6 | #105 |
| 8 | Gemini 2.5 Flash | 89.4 | #237 |
| 9 | GPT-5 | 87.6 | #91 |
| 10 | GPT-5.1 | 86 | #131 |
| 11 | GPT-4.1 | 82.1 | #240 |
| 12 | Llama 4 Maverick | 79.9 | #451 |
| 13 | GPT-4o | 79.5 | #333 |
| 14 | Claude Haiku 4.5 | 73.9 | #271 |
| 15 | DeepSeek V3.2 | 72 | #198 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=islamicmmlu-hadith-source-identification · How It Works · Data refreshed daily, snapshot 2026-10-11.