Can you please add in support for this github repo

[jjboi8708]

Github: https://github.com/Ilikepizza2/localspeech-AI

This uses OpenAI format api but I can't get it to work.

[richardr1126]

I'll try to. It seems like this repo's tts servers use a chat completion (openai_client.chat.completions.create()) with attached audio instead of the regularopenai_client.audio.speech.create(), which is what I currently utilize.

response = client.chat.completions.create(
    model="kokoro-tts",
    modalities=["text", "audio"],
    audio={
        "voice": "af_heart", 
        "format": "wav",
        "lang_code": "a",
        "speed": 1
    },
    messages=[
        {
            "role": "user",
            "content": [{"text": "This is the text to be converted to speech with a female voice.", "type": "text"}]
        }
    ]
)

[jjboi8708]

Yeah I noticed that why I tried to see if it could work. I think if you are able to implement this it would be great since its also tailored to mac too.