Evals for Every Language - Language umb — leaderboard
Metric: Average Score (%). Source: huggingface.co. 70 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 42.38 |
| 2 | Claude Opus 4.7 | 38.69 |
| 3 | Gemini 2.5 Pro | 37.89 |
| 4 | Gemini 3.1 Flash Lite | 35.58 |
| 5 | Claude Sonnet 4.5 | 33.54 |
| 6 | Claude Opus 4.8 | 33.52 |
| 7 | Grok 4.20 | 31.76 |
| 8 | Gemini 2.5 Flash | 31.62 |
| 9 | GPT-5.5 | 31.29 |
| 10 | GPT-5.4 | 30.47 |
| 11 | GPT-5.1 | 29.23 |
| 12 | GPT-4o | 28.97 |
| 13 | Command A | 28.36 |
| 14 | GPT-5 | 27.51 |
| 15 | Llama 3.1 70B Instruct | 27.2 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-umb · How the rankings work · Data refreshed daily, snapshot 2026-07-22.