Open Arabic LLM - Arabic MMLU History (Primary School): leaderboard
The "Arabic MMLU History (Primary School)" subtask of Open Arabic LLM. A task from the Open Arabic LLM Leaderboard, evaluating models on Arabic-language knowledge, reasoning and understanding across native and translated benchmarks.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around January 2027. 163 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 75.49 | #454 |
| 2 | EZO-Qwen2.5-32B-Instruct | 72.55 | #471 |
| 3 | Qwen2.5-32B-Instruct-CFT | 71.57 | #375 |
| 4 | Qwen 2.5 32B Instruct | 70.59 | #498 |
| 5 | Qwen 2.5 72B | 70.59 | #499 |
| 6 | Awqward2.5-32B-Instruct | 70.59 | #371 |
| 7 | tempmotacilla-cinerea-0308 | 70.59 | #395 |
| 8 | Yehia-7B-preview | 70.59 | #646 |
| 9 | gemma-3-27B-pt | 69.61 | #502 |
| 10 | Qwentile2.5-32B-Instruct | 69.61 | #339 |
| 11 | lambda-qwen2.5-14B-dpo-test | 68.63 | #481 |
| 12 | Qwen2.5-32B-Instruct-abliterated-v2 | 68.63 | #394 |
| 13 | Qwen3-8B-Base | 67.65 | #626 |
| 14 | ALLaM-7B-Instruct-preview | 67.65 | #718 |
| 15 | Fanar-1-9B | 67.65 | #778 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-history-primary-school · How It Works · Data refreshed daily, snapshot 2026-10-20.