Voxtral
By Mistral AI x OUTSCALE

Turn your audio and voice content into actionable data with Voxtral, offering advanced speech understanding for transcription. 

Voxtral Features

Voxtral is Mistral AI’s speech-to-text model that automatically transcribes conversations or audio files, generates summaries, and answers questions directly from voice content.

Transcription Mode

State-of-the-art transcription quality with automatic language detection. Compatible with 8 languages: English, Spanish, French, Portuguese, Hindi, German, Dutch, and Italian. 

Long-duration Audio Processing

The model can process long audio content for transcription and understanding. 

Multimodal Interactions

Audio only, text only, or both combined — Voxtral adapts to your needs. Ask questions about your voice files just as you would with a text document. 

Security and Sovereignty

Hosted on OUTSCALE’s Cloud infrastructure, certified ISO, HDS, and TISAX, and qualified SecNumCloud 3.2, Voxtral ensures the highest level of data and intellectual property protection.

Contact us now

Discover how our Cloud solutions can transform your projects.