European LLM Leaderboard - FLORES200: leaderboard
European LLM leaderboard translation score averaged over FLORES200 source and target directions across 21 European languages.
Metric: Average FLORES200 Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 49 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | EuroLLM-9B-Instruct | 25.6 | #1054 |
| 2 | Llama 3.1 70B Instruct | 24.2 | #561 |
| 3 | Mixtral 8x7B Instruct (v0.1) | 17.31 | #1073 |
| 4 | Llama 3.1 8B Instruct | 17 | #1035 |
| 5 | Mixtral 8x7B (v0.1) | 15.99 | #1083 |
| 6 | Llama 3 8B Instruct | 15.94 | #1129 |
| 7 | Llama 3.1 8B | 15.78 | #1151 |
| 8 | Llama 3 8B | 15.57 | #1242 |
| 9 | Mistral 7B Instruct (v0.3) | 13.26 | #1274 |
| 10 | vicuna-33B-v1.3 | 13.05 | #1347 |
| 11 | Mistral 7B Instruct (v0.2) | 12.38 | #1333 |
| 12 | vicuna-13B-v1.5 | 11.82 | #1392 |
| 13 | Mistral-7B-v0.1 | 11.78 | #1326 |
| 14 | Llama 2 13B Chat | 11.71 | #1465 |
| 15 | Mistral-7B-v0.3 | 11.43 | #1355 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=european-llm-leaderboard-flores200 · How It Works · Data refreshed daily, snapshot 2026-10-20.