Mistral releases the Voxtral 3 speech model family and Agent SDK
The story so far
Mistral AI released the Voxtral 3 voice model family and voice Agent SDK, including the voice-to-text model Voxtral Small Transcribe 3 and the text-to-speech model Voxtral Mini TTS 3. Its voice-to-text model performs excellently on the FLEURS 45 benchmark, achieving a Forced WER of 0.9%. The series models support use via Mistral's EU-hosted API or deployment on infrastructure, and provide real-time voice intelligence agent functionality. Mistral AI states that the voice-to-text service pricing starts at $0.0035 per minute.
AI-written from the reports · Updated 27 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- MarkTechPostMistral AI Releases Voxtral 3: Speech-to-Text, TTS and a Voice Agent SDK
Mistral AI has released the Voxtral 3 speech model family and the LiveKit-based speech intelligence agent SDK, including 3 transcription models and 1 text-to-speech model.
- The DecoderMistral AI launches Voxtral 3, a full voice agent stack that listens, thinks, and talks in real time
Mistral AI has released the real-time voice intelligence stack Voxtral 3, including a speech-to-text model (Voxtral Small Transcribe 3), a text-to-speech model (Voxtral Mini TTS 3), and a voice intelligence SDK based on LiveKit.
- Mistral AIVoxtral 3
Mistral AI has launched Voxtral 3, which includes an advanced voice model capable of real-time multi-step conversations in over 50 languages, along with a voice agent SDK. Its voice-to-text model, Voxtral Small Transcribe 3, achieved a Forced WER of 0.9% on the FLEURS 45 benchmark.