iuliandita/digarr

What it solves

Digarr is a self-hosted music discovery tool that helps users find new artists and albums based on their listening history, specific moods, or gaps in their existing music library. It automates the process of discovering music and integrating those findings into music management software or playlists.

How it works

The system connects to various listening sources (such as Spotify, Last.fm, and ListenBrainz) and library managers (like Plex or Jellyfin) to build a taste profile. It then uses AI providers (Anthropic, OpenAI, Gemini, or local models via Ollama) to generate personalized recommendations. Users can review these suggestions, listen to previews, and then approve them to be sent to acquisition tools like Lidarr or slskd, or export them as playlists to Spotify and other media servers.

Who it’s for

Music collectors and self-hosters who want an automated, AI-powered way to discover new music and sync those discoveries with their personal music libraries and media servers.

Highlights

• Flexible AI Integration: Supports major hosted AI providers and local models for privacy-conscious discovery.
• Broad Connectivity: Integrates with a wide array of services including Spotify, Deezer, MusicBrainz, Lidarr, and various media servers.
• Review Workflow: Allows users to preview, approve, reject, or permanently block suggestions before they are added to a library.
• Multi-user Support: Supports individual user queues, preferences, and OIDC/SSO authentication.