NTREX Translation - COMET — leaderboard
NTREX translation benchmark comparing English, Russian, and Chinese translation models across directions, with aggregate COMET, BLEU, and chrF views.
Metric: Average COMET. Source: huggingface.co. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | wmt19-en-ru | 0.85 |
| 2 | t5_translate_en_ru_zh_large_1024_v2 | 0.83 |
| 3 | t5_translate_en_ru_zh_large_1024 | 0.83 |
| 4 | madlad400-3b-mt | 0.83 |
| 5 | t5_translate_en_ru_zh_base_200 | 0.82 |
| 6 | opus-mt-ru-en | 0.82 |
| 7 | opus-mt-zh-en | 0.81 |
| 8 | mbart-large-ru-zh-ru-many-to-many-mmt | 0.81 |
| 9 | opus-mt-en-ru | 0.81 |
| 10 | t5_translate_en_ru_zh_small_1024 | 0.8 |
Interactive version: theaggregate.ai/benchmark?slug=ntrex-translation-comet · How the rankings work · Data refreshed daily, snapshot 2026-07-22.