Open Persian LLM - ARC Challenge: leaderboard
The "ARC Challenge" 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 | Gemma 3 27B (IT) | 90.6 | #523 |
| 2 | Llama 3.3 70B Instruct | 89.93 | #534 |
| 3 | QwQ 32B-Preview | 87.24 | #651 |
| 4 | Gemma 2 27B (IT) | 85.91 | #645 |
| 5 | Qwen 2.5 32B Instruct | 85.91 | #498 |
| 6 | Gemma 2 9B (IT) | 84.56 | #788 |
| 7 | Gemma 3 12B (IT) | 81.21 | #664 |
| 8 | aya-expanse-32B | 79.87 | #692 |
| 9 | aya-23-35B | 72.48 | #883 |
| 10 | aya-expanse-8B | 70.47 | #931 |
| 11 | Qwen 2.5 7B Instruct | 69.13 | #862 |
| 12 | Qwen 2 7B Instruct | 68.46 | #991 |
| 13 | aya-23-8B | 64.43 | #1299 |
| 14 | Gemma 3 4B (IT) | 63.76 | #992 |
| 15 | Llama 3.1 8B Instruct | 60.4 | #1035 |
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-arc-challenge · How It Works · Data refreshed daily, snapshot 2026-10-20.