Model dossier

Mistral: Voxtral Small Transcribe 3

Voxtral Small Transcribe 3 is Mistral's flagship speech-to-text model. It takes audio input and returns a transcript with timestamps and speaker diarization, and supports context biasing for domain-specific vocabulary. Suited...

ProviderMistral AI Start comparison

Context65,536

Max output52,428

Pricing viewTVP price

Identity

Mistral: Voxtral Small Transcribe 3

Model ID: mistralai/voxtral-small-transcribe-3-0
Canonical slug: mistralai/voxtral-small-transcribe-3-0-20261020
Knowledge cutoff: n/a
Created: 10/20/2026

Capabilities

audiotranscription

Modalities

audio->transcriptionaudiotranscription

Capacity

Context65,536

Max output52,428

TokenizerMistral