Models / FunAudioLLM / Fun CosyVoice3 0.5B

Fun CosyVoice3 0.5B

JSON →
audio

A lightweight 0.5B parameter text-to-speech model from FunAudioLLM, optimized for efficient and natural voice synthesis.

audiostreaming
Specs
Context & limits
context window
max output
Lifecycle
Dates
released2025-02-01
Resources