Open Arabic LLM - Madinah QA Average: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 163 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | free-evo-qwen72B-v0.8-re | 74.29 |
| 2 | Qwentile2.5-32B-Instruct | 73.6 |
| 3 | tempmotacilla-cinerea-0308 | 71.27 |
| 4 | Saka-14B | 70.01 |
| 5 | Gemma 3 27B (IT) | 69.43 |
| 6 | recoilme-gemma-2-9B-v0.4 | 67.43 |
| 7 | Qwen 2.5 72B | 65.02 |
| 8 | Qwen2.5-7B-Instruct-Uncensored | 64.23 |
| 9 | Qwen2.5-7B-Instruct-abliterated-v2 | 63.54 |
| 10 | Qwen2.5-7B-Instruct-abliterated | 63.01 |
| 11 | Qwen 2.5 7B Instruct | 62.73 |
| 12 | aya-23-35B | 62.49 |
| 13 | T.E-8.1 | 62.34 |
| 14 | Qwen 2 72B | 61.37 |
| 15 | Qwen 2.5 14B | 59.9 |
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-madinah-qa-average · How It Works · Data refreshed daily, snapshot 2026-09-05.