It would be really nice if we could consciously decouple from OpenAI and be able to use other compatible models and have it done on the server side for us. (and integrate this everywhere)
That is right now only OpenAI is supported. But I’d love to be able to run Ollama locally and use it with plexamp (or plex for recommendations) passing through to the server doing the recommendations for me. This would allow me to use open source models that aren’t expensive.
Bonus if we can have a built in SIRI/Google control + button tap what brings up voice chat or sends to PlexAmp so that I can do something like: “Siri, have plexamp play high energy hits. No back catalog. I’m in the mood for driving! Use my favorited music.” and it will just go and do it for me. “Siri have plexamp create a playlist from this mix. Name it Hard Driving”
