Mistral Launches Voxtral 3, a Model Suite for Building Voice AI Agents

At a glance

Mistral has released Voxtral 3, a new set of models for building AI voice agents that can listen, think, and speak. It includes a speech-to-text model, a text-to-speech model, and a software development kit (SDK) that links them together. It can be used through servers in Europe, and companies can also install it directly on their own servers.

Key points

1. 1

   The speech transcription model handles more than 50 languages, can tell who is speaking, and can help it better recognize specific words.

2. 2

   The speech synthesis model speaks naturally by changing emotion, speaking speed, and emphasis to suit the conversation.

3. 3

   Voice agents can look up information during a conversation and operate company systems to complete tasks over the phone or on the web.

4. 4

   The Pro Voice feature for experts can reproduce one person's speaking style and emotional expression in multiple languages using only 30 minutes of audio.

Why it matters

Services that handle reservations and consultations by voice are growing, so the ability to decide where data is processed matters a lot to businesses.