Open Arabic LLM - Arabic MMLU Social Science (Primary School) — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 163 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Ultiima-72B | 84.11 |
| 2 | Rombos-LLM-V2.5-Qwen-72B | 83.83 |
| 3 | Saka-14B | 82.27 |
| 4 | Qwen 2.5 72B Instruct | 81.28 |
| 5 | calme-2.2-qwen2.5-72B | 81.28 |
| 6 | gemma-3-27B-pt | 80.43 |
| 7 | calme-2.1-qwen2.5-72B | 80.14 |
| 8 | lambda-qwen2.5-14B-dpo-test | 80 |
| 9 | Llama 3.3 70B Instruct | 79.72 |
| 10 | tempmotacilla-cinerea-0308 | 79.72 |
| 11 | Qwen 2.5 14B | 79.57 |
| 12 | Qwen 2.5 72B | 79.43 |
| 13 | lambda-qwen2.5-32B-dpo-test | 79.29 |
| 14 | Qwen2.5-32B-Instruct-abliterated-v2 | 78.44 |
| 15 | Awqward2.5-32B-Instruct | 78.3 |
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-social-science-primary-school · How the rankings work · Data refreshed daily, snapshot 2026-07-22.