Sobre o projeto

AudioMuse‑AI analyzes audio directly via sonic fingerprinting and AI. It clusters similar tracks, creates playlists from natural‑language prompts, builds a 2D visual map, offers Song Paths for seamless transitions, generates listening‑habit playlists, and provides Song Alchemy with ADD/SUBTRACT controls. Search by mood, instrument, genre, or lyrics (72 languages) and identify songs from a 20‑second clip. It runs on multiple servers (Navidrome, Jellyfin, LMS, Lyrion, Emby, Plex) with duplicate detection. Deploy with Docker Compose, Podman, Kubernetes (Helm), or native apps for macOS (Apple Silicon), Linux (.deb/.rpm), Windows. Requires 4‑core CPU with AVX2, 8 GB RAM, NVMe SSD; GPU optional. Configured via browser wizard; PostgreSQL 15 needed. Community‑driven, not affiliated with audiomuse.ai.