Voxtral TTS

Open weights Active Audio gen voxtral-tts-2603

4B text-to-speech model with zero-shot voice cloning in 9 languages; weights under CC BY-NC 4.0.

Released
Mar 23, 2026
Context window
Not disclosed
Max output
Not disclosed
Parameters
4B
Input price
Not disclosed
Output price
Not disclosed

Prices are first-party list prices in USD at the standard tier. Last verified Oct 30, 2026.

Pricing

Launch priceMar 23, 2026
Current price since Mar 23, 2026

Specifications

Developer
Mistral AI
Model type
Audio gen
Architecture
Not disclosed
Parameters
4B
Input
Text Audio
Output
Audio
Knowledge cutoff
Not disclosed
License
Creative Commons Attribution-NonCommercial 4.0
API access
First-party API