Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Machine Translation research benchmarks – Page 2 · SOTA2 Research

Machine Translation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

English-to-German (En-De)

34.17BLEU Score

48

Mar 27, 2026

89.07Average Score

47

Apr 27, 2026

Real-world translation benchmarks En→Ja

CAT-Translate-3.3B

34.8Avg Score

46

Jun 23, 2026

FLORES-200 xx→en

Claude-3.5-Sonnet

97.68XCOMET

46

Apr 28, 2026

Flores-200 Romance group xx->en (test)

42.88BLEU

46

Feb 26, 2026

Flores-200 Romance group en->xx (test)

39.83BLEU

46

Feb 26, 2026

IWSLT en-de 2017 (test)

37.1BLEU

46

May 21, 2026

WMT En-Fr newstest 2014 (test)

RevCol-Transformer

43.4BLEU

46

Feb 26, 2026

EuroLLM-22B (old)

88.9Score

43

Feb 26, 2026

WMT En-De (newstest2014)

BiBERT Contextualized Embeddings + Stochastic Layer Selection + Fine-Tuning

31.26BLEU

43

Feb 26, 2026

Conversational (test)

77.3chrF2

42

Jun 30, 2026

64.4chrF2 Score

42

Jun 30, 2026

State vector (inn.)

75.8Accuracy

42

Feb 27, 2026

43.8BLEU

42

Feb 26, 2026

IT All-domain datastore (test)

3,270.84Throughput (Tok/Sec)

41

Feb 26, 2026

Flores-101 (test)

3,920Average Score

41

Mar 16, 2026

WMT English-French 2014 (test)

SOTA (Supervised)

45.6BLEU

41

Feb 27, 2026

31.59BLEU

40

May 12, 2026

GRPO for Seq2Seq MT

5.22Delta chrF++

39

May 18, 2026

FLORES-200 (devtest)

5.67Delta CHRF++

39

May 28, 2026

WMT En-Ro 2016 (test)

35.4BLEU

39

Feb 27, 2026

WMT14 English-French (newstest2014)

45.9BLEU

39

Feb 26, 2026

WMT16 German-English (test)

40.6BLEU

39

Feb 26, 2026

FLORES xx→en (test)

-65.05Score (de→en)

38

Feb 26, 2026

38.5BLEU Score

37

Apr 8, 2026

Page 2 of 58

Popular datasets

WMT FLORES IWSLT newstest Flores-200

Follow for updates

Join SOTA2 Contact