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

Text-to-Speech research benchmarks – Page 6 · SOTA2 Research

Text-to-Speech

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

LibriSpeech 2.2 hours subset (test-clean)

0.019WER

6

Feb 26, 2026

LibriSpeech (other)

Whisper + Llama 2 + MMS TTS

4.9CER

6

Feb 26, 2026

LibriSpeech clean

Whisper + Llama 2 + MMS TTS

4CER

6

Feb 26, 2026

LJ Speech (val)

2.5Time to 5% WER

6

Feb 26, 2026

ParaNet 100 sentences (test)

0Repeat Errors

6

Feb 26, 2026

Internal US English dataset (test)

4.582MOS

6

Feb 26, 2026

WordVoice 5A-en (test)

0.019Duration MAE

5

Jul 8, 2026

WordVoice zh 5A (test)

0.014Dur-MAE

5

Jul 8, 2026

Difficult-Word Benchmark French 500 words

0.28D-PER

5

Jul 7, 2026

Difficult-Word Benchmark German 500 words

0.3D-PER

5

Jul 7, 2026

Averaged Scenarios Full (test)

3.12WER

5

Jun 26, 2026

Chinese Dialects test-time scenario

3.13WER

5

Jun 26, 2026

Chinese Sketches test-time scenario

3.11WER

5

Jun 26, 2026

Slurred Speech test-time scenario

4.49WER

5

Jun 26, 2026

Children's Speech test-time scenario

3.01WER

5

Jun 26, 2026

Accented Speech test-time scenario

2.81WER

5

Jun 26, 2026

LJ Speech Dataset

4.05AutoPCP

5

Jun 25, 2026

LibriTTS (Unseen)

2.31WER

5

Jun 23, 2026

L2-ARCTIC Non-native

1.44WER

5

Jun 19, 2026

10.39WER

5

Jun 19, 2026

Mandarin Chinese (ZH) high-resource

4.35MOS

5

Jun 11, 2026

German (DE) high-resource

4.02MOS

5

Jun 11, 2026

English high-resource

4.61MOS

5

Jun 11, 2026

CV3-Eval Cross-lingual zh → en

4.32WER

5

Jun 8, 2026

CV3-Eval Cross-lingual en → zh

8.01WER

5

Jun 8, 2026

Page 6 of 13

Popular datasets

LibriSpeech Seed-TTS Multilingual LEMAS-TTS IndicTTS Database

Follow for updates

Join SOTA2 Contact