Open Arabic LLM - Arabic MMLU Economics (Middle School): leaderboard
The "Arabic MMLU Economics (Middle 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: Estimated already saturated. 163 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 82.76 | #454 |
| 2 | jais-adapted-70B-chat | 82.76 | #636 |
| 3 | lambda-qwen2.5-14B-dpo-test | 81.61 | #481 |
| 4 | Qwen2.5-32B-Instruct-abliterated-v2 | 81.61 | #394 |
| 5 | Saka-14B | 81.61 | #344 |
| 6 | AIC-1 | 81.61 | #231 |
| 7 | Llama 3.3 70B Instruct | 80.46 | #534 |
| 8 | aya-23-35B | 80.46 | #883 |
| 9 | AceGPT-v2-32B-Chat | 80.46 | #507 |
| 10 | Qwentile2.5-32B-Instruct | 80.46 | #339 |
| 11 | tempmotacilla-cinerea-0308 | 80.46 | #395 |
| 12 | Hala-9B | 80.46 | #504 |
| 13 | Gemma 3 27B (IT) | 79.31 | #523 |
| 14 | Qwen 2 7B Instruct | 79.31 | #991 |
| 15 | gemma-3-27B-pt | 79.31 | #502 |
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-economics-middle-school · How It Works · Data refreshed daily, snapshot 2026-10-20.