Voxtral

by Mistral AI

Mistral's speech-to-text and text-to-speech models plus a voice agent SDK, in 50+ languages.

Paid · from From $0.0035/minNew release

What people use it for

New: one real-time pipeline for listening, reasoning, tool calling and speaking, with EU hosting or self-deployment.

Pay-per-use API: transcription from $0.0035/min, Voxtral Speech 3 TTS $13 per 1M characters. Self-deployment also offered. Checked Oct 2026.

Prices change often. Always confirm on the official site.