Open Arabic LLM - Arabic MMLU HT High School Macroeconomics: leaderboard
The "Arabic MMLU HT High School Macroeconomics" 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: Estimated already saturated. 163 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 81.77 | #454 |
| 2 | Qwen 2 72B | 78.38 | #521 |
| 3 | AceGPT-v2-70B-Chat | 76.89 | #407 |
| 4 | Llama 3.3 70B Instruct | 76.55 | #534 |
| 5 | Qwen 2.5 72B | 76.37 | #499 |
| 6 | gemma-3-27B-pt | 75.41 | #502 |
| 7 | Gemma 3 27B (IT) | 74.75 | #523 |
| 8 | Qwen 2.5 32B Instruct | 74.09 | #498 |
| 9 | Qwen2.5-32B-Instruct-CFT | 73.95 | #375 |
| 10 | Awqward2.5-32B-Instruct | 73.82 | #371 |
| 11 | AIC-1 | 73.54 | #231 |
| 12 | Qwentile2.5-32B-Instruct | 73.4 | #339 |
| 13 | Qwen2.5-32B-Instruct-abliterated-v2 | 72.95 | #394 |
| 14 | EZO-Qwen2.5-32B-Instruct | 69.6 | #471 |
| 15 | AceGPT-v2-32B-Chat | 69.18 | #507 |
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-high-school-macroeconomics · How It Works · Data refreshed daily, snapshot 2026-10-20.