Open Arabic LLM - Arabic MMLU HT Business Ethics: leaderboard
The "Arabic MMLU HT Business Ethics" 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 | Llama 3.3 70B Instruct | 77 | #534 |
| 2 | Qwen 2.5 72B Instruct | 77 | #454 |
| 3 | AIC-1 | 75 | #231 |
| 4 | AceGPT-v2-70B-Chat | 73 | #407 |
| 5 | Qwen 2 72B | 73 | #521 |
| 6 | emirati-14B-v2 | 73 | #726 |
| 7 | gemma-3-27B-pt | 72 | #502 |
| 8 | Qwen 2.5 72B | 72 | #499 |
| 9 | Saka-14B | 72 | #344 |
| 10 | Gemma 2 27B (IT) | 71 | #645 |
| 11 | Awqward2.5-32B-Instruct | 71 | #371 |
| 12 | Qwen2.5-Lumen-14B | 71 | #477 |
| 13 | AceGPT-v2-32B-Chat | 70 | #507 |
| 14 | Qwentile2.5-32B-Instruct | 70 | #339 |
| 15 | Rombos-LLM-V2.6-Qwen-14B | 70 | #516 |
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-ht-business-ethics · How It Works · Data refreshed daily, snapshot 2026-10-20.