Audio / Text-to-Speech

MAI-Voice-1

Text-to-speech model from Microsoft Research & Azure · released Apr 2026

1025Speech Arena Elo (±12)
#79-86rank range of 105
2,494blind votes
4 voicesvoices in the arena
Apr 2026release

95% confidence interval: Elo 1013 – 1037. 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
Apr 2026 (month precision)
Release source
Artificial Analysis Speech Arena (month precision)

More TTS models from Microsoft Research & Azure