Mistral · 2026 · Voxtral
Voxtral TTS
Mistral’s 4B speech model (March 2026), with published streaming latency around 90 ms. CC / research-style licensing — check before you ship. Sample unavailable.
Known for
A frontier lab’s speech model that is not a TTS startup.
Sample unavailable. No redistributable clip was on file. Listen on the official site rather than a fake waveform.