Hello,
I currently am using Open AI’s tts-1 model for an app I am developing. One of the features is the AI types back and also speaks the typed response.
I am facing a latency issue where the AI is taking longer to start speaking than it is to start typing. I’ve tried the below solutions to get the AI TTS to start faster but no luck, let me know if there are any more solutions than what I have tried below:
- Streaming TTS
- Using faster gpt models
- Reduced max tokens per answer/reply