Machine Translation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
-48.8Error (en→de)
4
Feb 26, 2026
-3.7Score (en→de)
4
Feb 26, 2026
87.9faux-BLEU
4
Feb 26, 2026
89.1faux-BLEU
4
Feb 26, 2026
83.1faux-BLEU
4
Feb 26, 2026
88.8faux-BLEU
4
Feb 26, 2026
-50.3Score (de→en)
4
Feb 26, 2026
-5.5Score (de->en)
4
Feb 26, 2026
-14.4DE->EN Score
4
Feb 26, 2026
-51.4Score (de->en)
4
Feb 26, 2026
39.97BLEU
3
Jun 25, 2026
7.867Terminology
3
Jun 2, 2026
17BLEU
3
May 29, 2026
2.72Average Acceptance Length (tau)
3
May 15, 2026
85.27Mean DA Score
3
May 15, 2026
24.33sacreBLEU
3
Apr 21, 2026
15.39sacreBLEU
3
Apr 21, 2026
88.89Score
3
Apr 15, 2026
13.75BLEU
3
Apr 9, 2026
64.62BA Score
3
Apr 8, 2026
86.49COMET Score
3
Mar 27, 2026
87.05COMET Score
3
Mar 27, 2026
58Win-rate vs Gemini 3 Pro (preview)
3
Mar 27, 2026
38.033BLEU (linewise)
3
Mar 18, 2026
82.94COMET Score
3
Mar 16, 2026