Mistral AI launches Voxtral 3 speech models and voice agent toolkit

The release includes three transcription models, one text-to-speech model and a LiveKit-based voice agent SDK, according to MarkTechPost. Mistral says the offering supports more than 50 languages and real-time speech, with emotional text-to-speech and tool use. The company offers EU-hosted deployment or the option to deploy it independently.

• MarkTechPost reports Voxtral Small Transcribe 3 ranks third on Artificial Analysis with 2.1% AA-WER.

Covered by 2 publishers within 18 hours of the first report: 1 news report and 1 from the company.

• MarkTechPost: Mistral AI Releases Voxtral 3: Speech-to-Text, TTS and a Voice Agent SDK

Reporting1

MT MarkTechPost Mistral AI Releases Voxtral 3: Speech-to-Text, TTS and a Voice Agent SDK · Asif Razzaq 2:03 AM

From the company1

MIS Mistral AI Voxtral 3: Voice Agent SDK & Speech Models in 50+ Languages 8:00 AM

Related

TNW CERN and Mistral plan an AI foundation model for particle physics October 15

NV Mistral joins NVIDIA to develop AI workflows for engineering October 21

TRP Nvidia Launches Three Nemotron AI Models, Including Two for Local AI Agents October 15

TNW SandboxAQ launches tools to test AI agents and track what they cost October 19

Topics Mistral AI