Using Voxtral Mini Transcribe Realtime 3 on Mistral AI Studio

Implementation guide · Voxtral · MistralAI

Serverless

Mistral AI Studio exposes Voxtral Mini Transcribe Realtime 3 through model ID voxtral-mini-transcribe-realtime-3-0. Use the setup steps, sourced pricing, capabilities, and official provider links below to validate this route before deployment.

Last refreshed 2026-10-20. Next refresh: weekly.

Quick Start

  1. 1
    Create an account at Mistral AI Studio and generate an API key.
  2. 2
    Use the Mistral AI Studio SDK or REST API to call voxtral-mini-transcribe-realtime-3-0 — see the documentation for request format.
  3. 3
    You'll be billed $0.008/min. See full pricing.
API PortalDocumentationPricingModel Card

Code Examples

See Mistral AI Studio documentation for integration details.

Pricing on Mistral AI Studio

TypeRate
Audio minute$0.008/min

Capabilities

MultimodalAudio

About Voxtral Mini Transcribe Realtime 3

Voxtral Mini Transcribe Realtime 3 is Mistral AI's streaming speech-to-text model in the Voxtral 3 family, released October 20, 2026 as a Public Preview on the Mistral API. Built for voice agents, it streams transcripts over a single WebSocket session with configurable latency from 80 ms to 2,400 ms, built-in end-of-turn detection, context biasing and speaker diarization, and Mistral cites sub-500 ms latency in more than 35 languages. It is the listening stage of Mistral's Voxtral 3 voice agent SDK. It costs $0.008 per audio minute; weights are not published.

Full model details →

Model Specs

Released2026-10-20

More Models on Mistral AI Studio

Voxtral Mini Transcribe 2Voxtral Small 24B 2507Voxtral TTSVoxtral Mini Transcribe 3Voxtral Small Transcribe 3Voxtral TTS 3
All models on Mistral AI Studio →

Provider

Mistral AI Studio

Mistral AI

Paris, Île-de-France, France