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

Automatic Speech Recognition research benchmarks – Page 30 · SOTA2 Research

Automatic Speech Recognition

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Vietnamese (vi) low-resource ASR

Whisper Enc. + Sailor-7B

15.1WER

4

Feb 26, 2026

Thai (th) low-resource ASR

Whisper Enc. + Sailor-7B

0.161WER

4

Feb 26, 2026

Bangla Clean Speech v20 SLR53 Bhasha-bichitra (test)

BanglaRobustNet

12.3WER

4

Feb 26, 2026

CHiME-4 single-channel (clean)

WavLM (Weighted-sum)

80WER

4

Feb 26, 2026

Whisper-large-v3

1.47WER

4

Feb 26, 2026

Qwen2.5-Omni-7B

0.013WER

4

Feb 26, 2026

WenetSpeech-Yue (Short sentence)

6WER

4

Feb 26, 2026

WenetSpeech-Yue Long sentence

12.1WER

4

Feb 26, 2026

6.21CER

4

Feb 26, 2026

21CER

4

Feb 26, 2026

21.34CER

4

Feb 26, 2026

10.6CER

4

Feb 26, 2026

10.17CER

4

Feb 26, 2026

4.95CER

4

Feb 26, 2026

7.36CER

4

Feb 26, 2026

10.02CER

4

Feb 26, 2026

6.28CER

4

Feb 26, 2026

Hmong word+tone (test)

0.184CER

4

Feb 26, 2026

ML-SUPERB 10-min Few-shots 1.0

LID-CA-XLSR_dual^L

33.4ASR CER

4

Feb 26, 2026

Average of 6 public Chinese ASR sets

2.98WER

4

Feb 26, 2026

English Multi-accent (evaluation set)

11.26WER

4

Feb 26, 2026

English Multi-domain (val)

5.34WER

4

Feb 26, 2026

7.87WER (%)

4

Feb 26, 2026

TED-LIUM 2 (test)

ConMamba (All-in-One)*

6.8WER (HAT)

4

Feb 26, 2026

Callhome Eval2000 300-hour (test)

23.2WER

4

Feb 26, 2026

Page 30 of 38

Popular datasets

LibriSpeech FLEURS Common Voice AISHELL Multilingual ASR Corpus

Follow for updates

Join SOTA2 Contact