Open Arabic LLM - Arabic MMLU HT High School Biology: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 163 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 80.75 |
| 2 | Qwen 2 72B | 77.1 |
| 3 | Llama 3.3 70B Instruct | 75.37 |
| 4 | Qwen 2.5 72B | 74.98 |
| 5 | gemma-3-27B-pt | 74.35 |
| 6 | Gemma 3 27B (IT) | 73.64 |
| 7 | Awqward2.5-32B-Instruct | 72.65 |
| 8 | Qwen 2.5 32B Instruct | 72.62 |
| 9 | Qwen2.5-32B-Instruct-CFT | 72.49 |
| 10 | Qwentile2.5-32B-Instruct | 72.36 |
| 11 | Qwen2.5-32B-Instruct-abliterated-v2 | 71.75 |
| 12 | EZO-Qwen2.5-32B-Instruct | 68.37 |
| 13 | Gemma 2 27B (IT) | 68.21 |
| 14 | Saka-14B | 68.08 |
| 15 | free-evo-qwen72B-v0.8-re | 67.14 |
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-ht-high-school-biology · How It Works · Data refreshed daily, snapshot 2026-09-05.