Mistral's Voxtral 3 runs a full voice conversation in under a second
Mistral launches Voxtral 3, a family of speech models plus a voice agent SDK built on LiveKit that streams listening, reasoning, tool calling, and speaking as one pipeline, completing each turn in under a second. Voxtral Small Transcribe 3 handles 50+ languages with diarization and context biasing, beating Whisper, Scribe v2, and Gemini 3.5 Transcribe, while Voxtral Mini TTS 3 adds emotional, dialect-aware speech and voice cloning from 30 minutes of audio. Available via Mistral's EU-hosted API or self-deployment.
No stage waits for the one before it, so each turn completes in under one second, and the agent stops talking when the caller cuts in.