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 12 · SOTA2 Research

Text-to-Speech

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

2.65UTMOS

3

Feb 26, 2026

Seed-TTS-Eval ZH

2.92UTMOS

3

Feb 26, 2026

Seed-TTS-Eval EN

3.71UTMOS

3

Feb 26, 2026

Common Voice Romanian (1000 generated samples)

3.62WER

3

Feb 26, 2026

AISHELL-2 ZH (test)

5.8TTS CER

3

Feb 26, 2026

Librispeech (EN)

3.5WER

3

Feb 26, 2026

IndicTTS Database Rajasthani

4.1MOS

3

Feb 26, 2026

IndicTTS Database Gujarati

4.12MOS

3

Feb 26, 2026

IndicTTS Database Bengali

4.58MOS

3

Feb 26, 2026

0.131MAE

3

Feb 26, 2026

LibriTTS (460-hours clean sets)

3.87MOS

3

Feb 26, 2026

100 hard sentences (test)

Transformer TTS

0.15DFR (%)

3

Feb 26, 2026

LJSpeech low-resource setting

97Intelligibility Rate

3

Feb 26, 2026

50 particularly hard sentences (test)

0Repeats

3

Feb 26, 2026

Taiwan-localized 500-sentence (test)

328Sentences

2

Jul 8, 2026

Algerian Darija

4.12Naturalness

2

Jun 25, 2026

Melsolitica Malay (test)

Ground Truth (Melsolitica)

4.65MOS

2

Apr 15, 2026

LibriSpeech other (test)

15.1CER

2

Mar 11, 2026

Egyptian Arabic (evaluation)

18.8WER

2

Feb 26, 2026

LibriSpeech (test)

4.43MOS Naturalness

2

Feb 26, 2026

Bark official suno-ai implementation (test)

76.04MUSHRA Score

2

Feb 26, 2026

IndicTTS Database Manipuri

4.58MOS

2

Feb 26, 2026

IndicTTS Database Bodo

4.53MOS

2

Feb 26, 2026

IndicTTS Database Assamese

3.63MOS

2

Feb 26, 2026

Malaysian-TTS v2

Proposed Method (FS2 + Mimi)

11.6MCD

1

Apr 15, 2026

Page 12 of 13

Popular datasets

LibriSpeech Seed-TTS Multilingual LEMAS-TTS IndicTTS Database

Follow for updates

Join SOTA2 Contact