Hey everyone, have you heard any updates on the new voice mode tts-2 in the OpenAI API? It's strange that the OpenAI blog hasn't been updated with this release yet. It appears that the new voice mode has been rolled out to all users by now.
I'm curious if anyone has discovered how to utilize it in the API or if there have been any recent developments regarding it. When attempting to use tts-2 as the model, I encountered the error message 'model_not_found'. Here is a visual representation of the API versioning:
I also tested using gpt-4o as the model, but unfortunately, no success there either.
API Updates
While an API version is in the works, there is no specific timeline available for its release at the moment. It's great to see this progress, and I appreciate the insights provided by the community, such as @foxlabs.
Are there any updates on potential API modifications, or will it primarily involve a model update? Here is an illustration of API access as a ChatGPT Plus Subscriber:
I'm particularly interested to know if it will support voice-to-voice interactions or if there will be an option for text-to-speech functionality.
According to recent discussions, an API is in development, but it's still in the early stages of the development cycle, and details are limited at this point. One aspect that has been highlighted is the potential demand for a voice-to-voice endpoint, in addition to TTS (Text-to-Speech) and STT (Speech-to-Text) capabilities. The exact implementation, whether voice-to-voice will require two separate calls, remains uncertain.
This platform is powered by Discourse and is best experienced with JavaScript enabled.
