So Chatterbox is quite good at voice cloning, it’s new and rivals Eleven Labs in some ways. There is a project on Github for a Server serving this up but it’s not fully Open AI compatible (can’t use OpenAI TTS Device which is what I use to get my Kokoro Fast API into HA Voice pipeline).
Just something to keep an eye on, I had it clone my voice locally and it did a very decent job. Inference time was like 3 seconds for a standard message on an RTX 5000 Ada
But as I saw, Chatterbox do not support German language right? Its really a pity all new and good TTS solutions do not support German. Kokoro, Dia, Chatterbox…
Seem to be this way. Let’s hope they add German and other languages down the road. I assume since there is a lot of movement in the TTS sector investing deeply in one solution may not pay out.
Official reply form the Chatterbox discord “… those languages are supported with our last generation models - not based off chatterbox architecture. We are in the process of porting languages over to chatterbox architecure. More to come soon! Hope this helps!”