[FEATURE] More tts options

[razvanab]

Problem Statement

Using a custom voice

Proposed Solution

Adding more TTS options, something with voice cloning.

Alternatives Considered

Any alternative solutions or workarounds you've considered?
Piper will be a good choice too, but it is harder to train a custom voice.

Use Case

Describe how this feature would be used and who would benefit from it.
Using a custom voice for the Assistant

Additional Context

Any other context, mockups, or examples about the feature request.

[cmac86]

Hi @razvanab, thanks for the feature request. I've looked in to Chatterbox and QwenTTS as high quality options that also allow voice cloning. I just find the inference time too long with these models. Piper and Kokoro are both supported now.

Have you used any local TTS options that allow cloning and are fast enough at inference time for real-time conversation with an agent? I'm open to trying some others.

cmac