VibeVoice 1.5B

JSON →
vibevoice audio
audio

A 1.5 billion parameter text-to-speech model optimized for expressive voice generation.

streaming