Evals for Every Language - Language mg: leaderboard
Metric: Average Score (%). Source: huggingface.co. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 71.76 |
| 2 | Claude Opus 4.7 | 71.09 |
| 3 | Claude Opus 4.8 | 70.4 |
| 4 | GPT-5.5 | 70.14 |
| 5 | GPT-5 | 68.44 |
| 6 | Gemini 3.1 Flash Lite | 67.51 |
| 7 | GPT-5.1 | 66.41 |
| 8 | Claude Sonnet 4.6 | 66.4 |
| 9 | GPT-5.2 | 66.37 |
| 10 | Claude Sonnet 4.5 | 66.09 |
| 11 | GPT-5.3 | 66.05 |
| 12 | Claude Sonnet 4 | 65.74 |
| 13 | GPT-5.4 | 65.4 |
| 14 | Grok 4.20 | 63.81 |
| 15 | GPT-4o | 63.47 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-mg · How It Works · Data refreshed daily, snapshot 2026-09-05.