Wikibench

WMT24++

language

WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.

Leaderboard

Showing 20 of 20 results

  1. Nemotron 3 Super (120B A12B)
    86.7%
    i
  2. Nemotron 3 Nano (30B A3B)
    86.2%
    i
  3. Qwen3.7 Max
    85.8%
    i
  4. Qwen3.6 Plus
    84.3%
    i
  5. Qwen3.5-397B-A17B
    78.9%
    i
  6. Qwen3.5-122B-A10B
    78.3%
    i
  7. Qwen3.5-27B
    77.6%
    i
  8. Qwen3.5-35B-A3B
    76.3%
    i
  9. Qwen3.5-9B
    72.6%
    i
  10. Qwen3.5-4B
    66.6%
    i
  11. Gemma 3 27B
    53.4%
    i
  12. Gemma 3 12B
    51.6%
    i
  13. Gemma 3n E4B Instructed
    50.1%
    i
  14. Gemma 3n E4B Instructed LiteRT Preview
    50.1%
    i
  15. Gemma 3 4B
    46.8%
    i
  16. Qwen3.5-2B
    45.8%
    i
  17. Gemma 3n E2B Instructed
    42.7%
    i
  18. Gemma 3n E2B Instructed LiteRT (Preview)
    42.7%
    i
  19. Gemma 3 1B
    35.9%
    i
  20. Qwen3.5-0.8B
    27.2%
    i