Open Arabic LLM - Arabic MMLU HT High School Psychology — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 163 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Rombos-LLM-V2.5-Qwen-72B | 86.42 |
| 2 | calme-2.1-qwen2.5-72B | 86.06 |
| 3 | calme-2.2-qwen2.5-72B | 86.06 |
| 4 | Ultiima-72B | 86.06 |
| 5 | Qwen 2.5 72B Instruct | 84.95 |
| 6 | Gemma 3 27B (IT) | 83.12 |
| 7 | Qwen 2 72B | 82.57 |
| 8 | gemma-3-27B-pt | 80.55 |
| 9 | Gemma 2 27B (IT) | 80.18 |
| 10 | Qwen 2.5 72B | 79.45 |
| 11 | Llama 3.3 70B Instruct | 79.08 |
| 12 | Qwentile2.5-32B-Instruct | 78.35 |
| 13 | lambda-qwen2.5-32B-dpo-test | 77.98 |
| 14 | Awqward2.5-32B-Instruct | 77.43 |
| 15 | Qwen 2.5 32B Instruct | 76.7 |
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-ht-high-school-psychology · How the rankings work · Data refreshed daily, snapshot 2026-07-22.