Open Arabic LLM - Arabic MMLU Islamic Studies (High School): leaderboard
The "Arabic MMLU Islamic Studies (High 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 | 76.65 | #454 |
| 2 | Llama 3.3 70B Instruct | 75.75 | #534 |
| 3 | EZO-Qwen2.5-32B-Instruct | 75.15 | #471 |
| 4 | Awqward2.5-32B-Instruct | 74.85 | #371 |
| 5 | Qwen 2.5 32B Instruct | 74.25 | #498 |
| 6 | Qwen2.5-32B-Instruct-CFT | 74.25 | #375 |
| 7 | Qwen2.5-32B-Instruct-abliterated-v2 | 73.35 | #394 |
| 8 | gemma-3-27B-pt | 72.46 | #502 |
| 9 | Qwen 2 72B | 72.46 | #521 |
| 10 | Qwentile2.5-32B-Instruct | 72.16 | #339 |
| 11 | AceGPT-v2-32B-Chat | 71.86 | #507 |
| 12 | Qwen 2.5 72B | 71.56 | #499 |
| 13 | lambda-qwen2.5-14B-dpo-test | 71.56 | #481 |
| 14 | Saka-14B | 71.56 | #344 |
| 15 | Gemma 3 27B (IT) | 71.26 | #523 |
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-islamic-studies-high-school · How It Works · Data refreshed daily, snapshot 2026-10-20.