[feature] TTS support request

[PylotLight]

Is your feature request related to a problem? Please describe.
Adding support for extra TTS models/backends?

Describe the solution you'd like
Add support for these models:
https://github.com/yl4579/StyleTTS2
https://github.com/myshell-ai/OpenVoice
https://github.com/collabora/WhisperSpeech

Describe alternatives you've considered
Perhaps I don't understand backend vs model so not sure if these are already supported or not.

Additional context
N/A

[localai-bot]

Hello PylotLight,

Thank you for submitting your feature request for TTS support in LocalAI. I am a bot, an experiment of @mudler, and I will try to assist you with your issue.

From your request, it seems you would like to see support for additional Text-to-Speech (TTS) models in LocalAI. You have mentioned three models:

1. StyleTTS2: https://github.com/yl4579/StyleTTS2
2. OpenVoice: https://github.com/myshell-ai/OpenVoice
3. WhisperSpeech: https://github.com/collabora/WhisperSpeech

To clarify, LocalAI is a versatile tool that supports various models, and it would be great to have these TTS models integrated in case they are not yet supported. If you are unsure whether a specific model is already supported, you can check the official LocalAI documentation or the models list in the LocalAI GitHub repository.

If you are interested in contributing to the LocalAI project and providing support for these TTS models, you can follow the contribution guidelines specified in the LocalAI repository.

In case you do not have the technical means to contribute, you can simply upvote this issue and tag @mudler or one of the other maintainers to review and possibly work on this enhancement.

Let me know if you have any more questions or need further assistance.

[Crystora]

@mudler Is this feature needed in LocalAI project? If so, pls let me know. I can solve this

[localai-bot]

LocalAI now ships a broad TTS lineup (Kokoro, Chatterbox, VibeVoice, F5-TTS, and more). The three specific projects here (StyleTTS2, OpenVoice, WhisperSpeech) are not individually integrated. Shall we close this and open a focused request if you still need one of those specifically?

[localai-bot]

Closing as completed. TTS is now a first-class feature in LocalAI: there is an OpenAI-compatible /v1/audio/speech endpoint plus a /tts endpoint, and many TTS backends are available in the gallery (piper, kokoro, coqui, chatterbox, outetts, vibevoice, qwentts, and more). See the text-to-speech docs. Thanks for the original request that helped kick this off!