Evals for Every Language - Classification: leaderboard
LanguageBench Classification accuracy averaged across available languages and scaled to 0-100. Translation uses spBLEU only, without mixing chrF; higher is better.
Metric: Average Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 70 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 98.87 | #152 |
| 2 | Claude Sonnet 4.5 | 98.87 | #142 |
| 3 | Gemini 3.1 Pro (Preview) | 98.25 | #58 |
| 4 | Claude Sonnet 4 | 97.84 | #199 |
| 5 | GPT-4o | 96.86 | #343 |
| 6 | GPT-5 | 96.75 | #98 |
| 7 | Claude Opus 4.7 | 95.98 | #49 |
| 8 | GPT-5.1 | 95.98 | #135 |
| 9 | GPT-5.3 | 95.57 | #103 |
| 10 | Gemini 3.1 Flash Lite | 95.5 | #191 |
| 11 | Gemma 3 27B (IT) | 95.41 | #523 |
| 12 | Command A | 95.36 | #508 |
| 13 | Nova Pro (v1) | 95.21 | #497 |
| 14 | Mistral Medium 3.1 | 94.48 | #408 |
| 15 | GPT-3.5 Turbo | 92.22 | #865 |
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-classification · How It Works · Data refreshed daily, snapshot 2026-10-20.