Open Arabic LLM - Alghafa Multiple Choice Rating Sentiment Task: leaderboard
The "Alghafa Multiple Choice Rating Sentiment Task" 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 2029. 163 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | c4ai-command-r7B-arabic-02-2025 | 58.33 | #732 |
| 2 | Llama 3.3 70B Instruct | 58.3 | #534 |
| 3 | free-evo-qwen72B-v0.8-re | 57.83 | #435 |
| 4 | Qwen 2 72B | 57.66 | #521 |
| 5 | aya-23-35B | 57.38 | #883 |
| 6 | jais-adapted-70B-chat | 57.05 | #636 |
| 7 | Hala-9B | 56.31 | #504 |
| 8 | AceGPT-v2-70B-Chat | 56.21 | #407 |
| 9 | Qwen 2.5 72B | 55.26 | #499 |
| 10 | Qwen 2.5 14B | 55.01 | #725 |
| 11 | aya-expanse-32B | 54.98 | #692 |
| 12 | Qwen 2.5 32B Instruct | 54.96 | #498 |
| 13 | Qwen2.5-32B-Instruct-CFT | 54.96 | #375 |
| 14 | Qwen 2.5 7B | 54.95 | #922 |
| 15 | Qwentile2.5-32B-Instruct | 54.93 | #339 |
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-alghafa-multiple-choice-rating-sentiment-task · How It Works · Data refreshed daily, snapshot 2026-10-20.