Generative AI Microsoft releases VibeVoice-ASR: a unified speech-to-text model designed to handle up to 60 minutes of long audio in a single pass by January 22, 2026 January 22, 2026 Read more
