audiomuse-ai

AudioMuse-AI analyzes the actual audio content of your music library using sonic fingerprinting and AI models to enable features like automatic clustering of sonically similar tracks, instant playlist generation from natural language descriptions (e.g., "high-tempo, low-energy music"), a visual 2D Music Map of your collection, Song Paths that bridge two tracks seamlessly, Sonic Fingerprint playlists based on listening habits, Song Alchemy for vibe-based curation with ADD/SUBTRACT controls, text search by mood/instrument/genre, lyrics search across 72 languages, and Search by Recording (identify a song from a 20-second clip or find alternate recordings). It supports multiple music servers simultaneously (Navidrome, Jellyfin, LMS, Lyrion, Emby, Plex) with built-in duplicate detection so each track is analyzed once and results are shared. Deployment options include Docker Compose, Podman, Kubernetes (Helm chart available, AMD64/ARM64), and native applications for macOS (Apple Silicon), Linux (x86_64/ARM64 .deb/.rpm), and Windows (x86_64). Minimum hardware: 4-core CPU with AVX2 (Intel 2015+ or ARM), 8 GB RAM, NVMe SSD. GPU acceleration (NVIDIA, experimental ARM DGX Spark) is optional for faster analysis. Configuration is managed via a browser-based Setup Wizard; PostgreSQL 15 is required. The project is community-driven and not affiliated with audiomuse.ai.