Open Persian LLM - AUT Multiple Choice Persian: leaderboard
The "AUT Multiple Choice Persian" subtask of Open Persian LLM. A task from the Open Persian LLM Leaderboard, evaluating models on Persian-language knowledge, reasoning and understanding.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 31 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.3 70B Instruct | 71.4 | #534 |
| 2 | Gemma 3 27B (IT) | 64.3 | #523 |
| 3 | Gemma 2 27B (IT) | 63.12 | #645 |
| 4 | Gemma 2 9B (IT) | 62.33 | #788 |
| 5 | aya-expanse-32B | 62.33 | #692 |
| 6 | Qwen 2.5 32B Instruct | 61.34 | #498 |
| 7 | aya-23-35B | 61.14 | #883 |
| 8 | QwQ 32B-Preview | 60.15 | #651 |
| 9 | aya-expanse-8B | 58.38 | #931 |
| 10 | Gemma 3 12B (IT) | 57 | #664 |
| 11 | Llama 3.1 8B Instruct | 54.24 | #1035 |
| 12 | Llama 3 8B Instruct | 53.85 | #1129 |
| 13 | Llama 3.1 8B | 53.45 | #1151 |
| 14 | aya-23-8B | 52.86 | #1299 |
| 15 | Qwen 2.5 7B Instruct | 52.66 | #862 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-persian-llm-aut-multiple-choice-persian · How It Works · Data refreshed daily, snapshot 2026-10-20.