I am very excited and grateful for the launch of advanced voice mode, and also the new voices.
But regarding the API, is there any plan to expand its audio capabilities, like adding more voices or to choose the language to be spoken?
1 Like
I’ve created a topic on the same thing, no replies there. I’ve looked into the source code of the API. Nothing there also. I’ve tried using “tts-2” on the voice endpoint (model doesn’t exist error) and “gpt-4o”, no luck there also. I guess we’ll just have to be patient or accept this won’t be releasing to the API at all. No communication on all social media channels has been done on this, so we are effectively on the dark for now (as far as I know).
There is an API planned, but there are no timelines as yet for when it will be released.
2 Likes
I see. Well, let’s hope it won’t take too long.
But considering there is already a realtime API being released, maybe it won’t take much longer until we get some new voices.
Would be really nice to have. I feel like Open-AI might be coupling the release of more voices to custom voices. We just want more options! Don’t need a custom.
2 Likes
The election is over. Time to do this. If you’re worried about safety…we already have elevenlabs…
See October 30: https://platform.openai.com/docs/changelog
But not the TTS api yet, right?
You are correct. This set of updated voices is only those generated by gpt-4o, not the transcriptions audio endpoint employing a different model.