I just want to share my free and opensource GitHub repo, AudioMuse-AI: https://github.com/NeptuneHub/AudioMuse-AI

It is able to do sonic analysis locally on your homelab thanks to Librosa and Tensorflow and the thanks of this analysis can find sonic similar song, useful for instant mix, playlist creation or song path.

Thanks to API it works with Jellyfin and all the open subsonic API music server like Navidrome or Lightweight Music Server (LMS).

If you like please share your feedback on GitHub and leave a star on the repository.