Evals for Every Language - Urdu (ur): leaderboard
Average score across evaluation tasks in Urdu (ur), from Evals for Every Language.
Metric: Average Score (%). Source: huggingface.co. Saturation forecast: Around January 2027. 67 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Flash Lite | 76.82 | #191 |
| 2 | Claude Opus 4.7 | 76.52 | #49 |
| 3 | Claude Sonnet 4.5 | 76.44 | #142 |
| 4 | Gemini 3.1 Pro (Preview) | 75.24 | #58 |
| 5 | GPT-5 | 74.79 | #98 |
| 6 | Gemini 2.5 Pro | 74.76 | #152 |
| 7 | GPT-5.3 | 73.72 | #103 |
| 8 | Grok 4.20 | 73.7 | #175 |
| 9 | Claude Sonnet 4 | 73.59 | #199 |
| 10 | Claude Sonnet 4.6 | 73.58 | #90 |
| 11 | Claude Opus 4.8 | 73.5 | #37 |
| 12 | GPT-5.1 | 73.2 | #135 |
| 13 | GPT-5.2 | 73.01 | #110 |
| 14 | DeepSeek V3.1 | 72.3 | #264 |
| 15 | Llama 3 70B Instruct | 71.07 | #634 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-ur · How It Works · Data refreshed daily, snapshot 2026-10-20.