Voxtral Small / Mini

Voxtral Small (also called Voxtral Mini) is Mistral AI's compact speech understanding model, designed to process and analyze audio inputs including speech, conversations, and other audio content. It combines audio encoding with language modeling to enable speech-to-text, audio analysis, and voice-based interaction.

The model handles multiple languages and accents, providing transcription, translation, and audio content understanding. It can process long audio recordings and supports real-time streaming for live applications.

Voxtral Small is optimized for efficient audio processing, suitable for voice assistants, meeting transcription, podcast analysis, and other audio understanding tasks.