Evals for Every Language: leaderboard
Multilingual LLM evaluation across 100+ languages: tests classification, generation, and comprehension tasks to measure language coverage and quality.
Metric: Average Score (%). Source: huggingface.co. Status: years away from saturation. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 69.19 |
| 2 | Claude Opus 4.7 | 66.95 |
| 3 | Claude Opus 4.8 | 66.27 |
| 4 | Gemini 3.1 Flash Lite | 65.25 |
| 5 | GPT-5.5 | 65.09 |
| 6 | Claude Sonnet 4.6 | 64.44 |
| 7 | Claude Sonnet 4.5 | 64.39 |
| 8 | Gemini 2.5 Pro | 64.15 |
| 9 | GPT-5.3 | 63.51 |
| 10 | Grok 4.20 | 63.06 |
| 11 | GPT-5.4 | 62.95 |
| 12 | GPT-5 | 62.87 |
| 13 | Qwen 3.6 Plus | 62.49 |
| 14 | Claude Sonnet 4 | 62.43 |
| 15 | GPT-5.2 | 62.37 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language · How It Works · Data refreshed daily, snapshot 2026-09-05.