KittenTTS v0.8: three new models, smallest under 25 MB
2mon 1h ago by lemdro.id/u/mudkip in localllama@sh.itjust.works from github.com
Agree. I think the developers stated they added cartoon voices on purpose to demonstrate expressiveness.
The demo video sounds really good. In testing, it is very noisy and poor quality even with the 80M model. It is not all that fast, not instant like I would expect. It's okay for memory constrained environments, but without voice cloning, may as well stick with whatever built-in TTS your OS has.