Mistral's new speech model can run on a smartwatch or a smartphone.
news
Mistral releases a new open-source model for speech generation
26. March 2026
