Oh right, yeah I've done things like this (phone calls to ChatGPT) or the openwebui Whisper -> LLM -> TTS setup. I thought there might be something more than this by now