Open Arabic LLM - Arabic MMLU Math (Primary School) — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 163 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Rombos-LLM-V2.5-Qwen-72B | 80.2 |
| 2 | Ultiima-72B | 80.2 |
| 3 | Gemma 3 27B (IT) | 79.46 |
| 4 | gemma-3-27B-pt | 79.22 |
| 5 | lambda-qwen2.5-32B-dpo-test | 78.97 |
| 6 | Qwen2.5-32B-Instruct-abliterated-v2 | 78.24 |
| 7 | Awqward2.5-32B-Instruct | 78 |
| 8 | Qwen 2.5 32B Instruct | 77.26 |
| 9 | Qwen2.5-32B-Instruct-CFT | 77.26 |
| 10 | Qwentile2.5-32B-Instruct | 75.06 |
| 11 | Qwen 2.5 72B Instruct | 74.82 |
| 12 | tempmotacilla-cinerea-0308 | 74.82 |
| 13 | Qwen 2.5 14B | 74.33 |
| 14 | Saka-14B | 74.08 |
| 15 | Qwen3-8B-Base | 73.59 |
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-math-primary-school · How the rankings work · Data refreshed daily, snapshot 2026-07-22.