Evals for Every Language - Language ars — leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4 | 53.43 |
| 2 | Gemini 2.5 Flash Lite | 53.25 |
| 3 | Command A | 52.85 |
| 4 | Gemini 2.5 Flash | 52.45 |
| 5 | Claude Sonnet 4.5 | 52.07 |
| 6 | Gemini 2.5 Pro | 51.41 |
| 7 | GPT-4o | 50.57 |
| 8 | Llama 4 Maverick | 50.51 |
| 9 | Qwen 3 30B A3B 2507 Instruct | 50.46 |
| 10 | Gemini 3.1 Pro (Preview) | 50.3 |
| 11 | Nova Pro (v1) | 50.29 |
| 12 | Claude Opus 4.7 | 49.83 |
| 13 | GPT-5 Nano | 49.79 |
| 14 | Gemma 3 27B (IT) | 49.33 |
| 15 | Gemini 3.1 Flash Lite | 49.28 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-ars · How the rankings work · Data refreshed daily, snapshot 2026-07-22.