FLORES European Languages Leaderboard - BLEU: leaderboard

FLORES European Languages BLEU leaderboard averages translation BLEU across published European-language pairs.

Metric: Average BLEU. Source: huggingface.co. Saturation forecast: Estimated already saturated. 20 models tracked.

Top models

#ModelScoreOverall rank
1EuroLLM-9B-Instruct20.61#1054
2Gemma 3 12B (IT)19.61#664
3Phi-415.58#706
4Qwen 3 14B15.37#537
5Mistral Nemo Instruct (2407)14.35#936
6cogito-v1-preview-qwen-14B13.74
7Qwen 3 8B13.48#674
8Qwen 2.5 14B Instruct13.24#647
9Qwen 2.5 7B Instruct11.34#862
10Bielik-11B-v211.25#1329
11Mistral 7B Instruct (v0.2)10.98#1333
12Bielik-7B-Instruct-v0.15.74#1292

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=flores-european-languages-leaderboard-bleu · How It Works · Data refreshed daily, snapshot 2026-10-20.