Audio / Text-to-Speech

WaveNet

Text-to-speech model from Google DeepMind · released Sept 2016

916Speech Arena Elo (±12)
#95-97rank range of 105
6,466blind votes
$4.00 / 1M charsAPI price
8 voicesvoices in the arena
Sept 2016release

95% confidence interval: Elo 904 – 928. Models whose intervals overlap share a rank range.

How this is measured

Speech Arena (Text-to-Speech)

Elo rating (higher is better)

Models are ranked with an Elo system derived from blind user votes: listeners hear two speech samples generated from the same text and pick the one that sounds more natural. The provider-voice leaderboard uses each provider's own native voices. Each result is reported with a 95% interval and the number of votes (samples).

Release & sources

Release
Sept 2016 (month precision)
Release source
Artificial Analysis Speech Arena (month precision)

More TTS models from Google DeepMind