Music Fetch
One question — what is this track? — answered for inputs that are usually hopeless: noisy web video, whole playlists, local files. One engine behind a CLI, a local HTTP API, a terminal UI, and a native macOS app.
A local-first, desktop-first music recognition tool. Ingestion follows whatever yt-dlp supports — YouTube, Shorts, Instagram, TikTok, Vimeo — plus anything already on disk. The engine scans overlapping windows from the original mix and, when available, a separated instrumental stem, so a voice-over or crowd noise doesn't end the search. Vocal/instrumental separation cleans noisy audio before matching.
Short clips use aggressive early-stop logic; long videos and mixes switch to a clustered mode with request budgets, excerpt caching, and repeated hits fused into timeline segments. Detached jobs can be submitted, watched, and inspected down to per-provider metrics.
The same engine is exposed four ways: a CLI (analyze, jobs), a local HTTP API, a terminal UI, and a native SwiftUI macOS app. The provider chain starts local and free — a chromaprint fingerprint catalog of your own music, then vibra (unofficial Shazam matching) — with hosted services (AudD, ACRCloud) only as an opt-in fallback when you bring your own credentials. Public at github.com/x1f4r/music-fetch.