MA
AI News: 24 Things You Need To Know This WeekMatt Wolfe23 Oct 2026
Mistral's voice agent model that handles multiple speakers with real-time diarization, interruptions, and tool calls, shown in a demo video.
Mistral's voice model for speech agents, with real-time diarization, interruption handling and tool calls.
Updated EU status checked Last shown
Voxtral 3 at a glance
- What is Voxtral 3?
- Voxtral 3 is a model from Mistral. Mistral's voice model for speech agents, with real-time diarization, interruption handling and tool calls.
- Who makes Voxtral 3?
- Voxtral 3 is made by Mistral.
- Does Voxtral 3 work in the EU?
- Available in the EU. Mistral is a France-based vendor offering EU pricing (EUR), in-region inference and on-prem/VPC deployment, and cites European customers like the European Patent Office. No Voxtral 3-specific availability page was provided. Last verified 14 Nov 2026.
- How much does Voxtral 3 cost?
- Pricing unknown. No Voxtral 3 pricing found in the provided pages. Mistral's Vibe plans: Free, Pro $14.99/mo, Team $24.99/user/mo, Enterprise custom.
- Where was Voxtral 3 shown?
- In 1 creator video by Matt Wolfe. Most recently by Matt Wolfe on 23 Oct 2026, at 20:50 in “AI News: 24 Things You Need To Know This Week”.
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
About
Voxtral 3 is a speech and voice-agent model from Mistral, the successor to the earlier Voxtral line. Based on a creator demo, it handles conversations with multiple speakers, including real-time speaker diarization, interruptions (barge-in) and tool calls during a conversation. The official pages provided do not include a dedicated Voxtral 3 model page, so technical details are limited.
Mistral offers its models via its own cloud with in-region inference, in a customer VPC, or on-prem, and sells voice features through Le Chat/Vibe (voice mode with transcription, questions and summaries). Customers such as the European Patent Office are cited as using Mistral for transcription. Exact model size, languages, latency, API pricing and license for Voxtral 3 could not be confirmed from the provided material.
Key features
- Real-time speaker diarization for multi-speaker audio
- Handles user interruptions in voice conversations
- Tool calling from within voice agents
- Deployable via Mistral cloud, VPC or on-prem
1 moment
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Keep exploring
Similar AI tools
Closest neighbours by meaning, not just by category.
MA
MAI-Voice-2.1
Model · Microsoft
Microsoft AI's text-to-speech voice model with expressive output across multiple languages.
UnverifiedVoice & Audio
Release date n/a
MA
MAI-Transcribe-2
Model · Microsoft
Microsoft AI's second-generation speech-to-text model for transcribing noisy, domain-specific audio.
UnverifiedVoice & Audio
3 Sept 2026
MS
Microsoft Streaming Transcription Model
Model · Microsoft
Microsoft AI's streaming speech-to-text model for near real-time transcription, succeeded by MAI Transcribe 2.
UnverifiedVoice & Audio
Release date n/a
G3
Gemini 3.5 Transcribe
Model · Google
Google's speech-to-text model with verbatim and smart modes, speaker detection, word timestamps and a live streaming variant.
UnverifiedVoice & Audio
Release date n/aAPI
MF
MAI-Voice-2 Flash
Model · Microsoft
Microsoft AI's low-latency text-to-speech model in the MAI model family, designed for long-form expressive speech.
UnverifiedVoice & Audio
Release date n/a
G3
Gemini 3.8 Live
Model · Google
Google's real-time voice model with an Extended Thinking variant, available in the Gemini app, AI Studio and Search Live.
UnverifiedVoice & Audio
Release date n/aAPI