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 33 · SOTA2 Research

Automatic Speech Recognition

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

WenetSpeech n (Net)

6.46WER (%)

3

Mar 12, 2026

Bengali-Loop (test)

27WER

3

Mar 6, 2026

IndicVoices Others avg

23.2WER

3

Mar 4, 2026

SAP Overall 2 (test)

8.1WER

3

Feb 26, 2026

17.8MER

3

Feb 26, 2026

AISHELL-2 1.0 (test)

6CER (Offline, Rescoring)

3

Feb 26, 2026

2.1WER

3

Feb 26, 2026

Qwen2.5-Omni-7B

1.3WER

3

Feb 26, 2026

11.43WER

3

Feb 26, 2026

CommonVoice (dev)

7.2WER

3

Feb 26, 2026

Kimi-Audio-Instruct

0.62WER

3

Feb 26, 2026

Indonesian Synthetic Clean Speech (test)

Whisper (Stutter only)

6.4WER

3

Feb 26, 2026

Indonesian Synthetic Stuttered Speech (test)

Whisper (Stutter only)

12.6WER

3

Feb 26, 2026

ASR (evaluation set)

1.6WER

3

Feb 26, 2026

Salt multispeaker eng

parakeet-rnnt-1.1b

3.1WER

3

Feb 26, 2026

parakeet-tdt-1.1b

15.8WER

3

Feb 26, 2026

parakeet-rnnt-1.1b

26.1WER

3

Feb 26, 2026

AISHELL-2 ZH (test)

7.2CER

3

Feb 26, 2026

KeSpeech (test)

3.56CER

3

Feb 26, 2026

Europarl-ASR EN Guest (test)

0.412WER

3

Feb 26, 2026

Fleurs zh (test)

7WER

3

Feb 26, 2026

Video Average of 7 subsets

0.027WER

3

Feb 26, 2026

1.94WER

3

Feb 26, 2026

Multilingual Multi-domain (evaluation set)

12.16WER

3

Feb 26, 2026

LibriCSS (test)

Conformer-transducer (IHM+)

0.25Ins. Error

3

Feb 26, 2026

Page 33 of 38

Popular datasets

LibriSpeech FLEURS Common Voice AISHELL Multilingual ASR Corpus

Follow for updates

Join SOTA2 Contact