Developers building transcription or meeting notes who want low per-minute prices; the realtime model also comes with open weights.
Kenning's take
What it does
Voxtral Transcribe 2 is a pair of speech-to-text models from Mistral AI. Voxtral Mini Transcribe V2 handles recorded audio with speaker labels, word timestamps and custom vocabulary, while Voxtral Realtime transcribes live audio with delay configurable down to under 200 milliseconds. Both work in 13 languages through Mistral's API, Mistral Studio and Le Chat, and Voxtral Realtime's weights are open under Apache 2.0.
- Speaker diarization and word-level timestamps
- Context biasing with up to 100 words or phrases
- Live transcription with sub-200ms latency
- 13 languages
- Voxtral Realtime open weights under Apache 2.0
Who it suits. Developers building transcription, subtitling, meeting notes or voice agents.
Key facts
- Category
- AI models
- Part of
- Le Chat (Vibe by Mistral)
- Made by
- Mistral AI
- Released
- 4 February 2026
- How to get it
- Mistral API, Mistral Studio audio playground and Le Chat; Realtime weights on Hugging Face
- Works on
- API, Web, open-weights
- In the UK
- Yes
- Free plan
- Voxtral Realtime open weights under Apache 2.0
- API price
- Voxtral Mini Transcribe V2 $0.003 per audio minute; Voxtral Realtime $0.006 per audio minute
FAQ
When was Voxtral Transcribe 2 released?
Mistral AI released it on 4 February 2026.
How much does Voxtral Transcribe 2 cost?
$0.003 per minute for batch transcription and $0.006 per minute for Voxtral Realtime on Mistral's API.
Is Voxtral Transcribe 2 open source?
Voxtral Realtime's weights are on Hugging Face under Apache 2.0. Voxtral Mini Transcribe V2 is API-only.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
Devstral 2
Mistral's open-weight coding model family, now retired from the API in favour of Mistral Medium 3.5
Mistral 3
Mistral's December 2025 family of open models, from the small Ministral 3 sizes up to Mistral Large 3
Mistral Small 4
An open Mistral model that handles chat, reasoning, coding and images in one 119B mixture-of-experts model
Mistral OCR 3
Mistral's December 2025 document reading model that pulls text, tables and images out of PDFs and scans