Rekordbox Library AI Companion
Your library, finished. Crate Control fills in everything rekordbox is missing — genre, artist, artwork, BPM, key, moods, vocal components, a 1–10 energy score and auto cue points — and builds playlists of tracks that play well together, using AI that runs entirely on your Mac. You review every proposal. Nothing is written until you press Apply.
Coming soon for macOS · one-time purchase, no subscription
Your library at a glance: what's complete, what the AI has covered, and every track with its BPM, key, energy, moods and components.
Auto-tag genre, energy and key · set cue points on the beat grid · build playlists of similar tracks
A ladder of matching tiers, not one search: files bought from Beatport match by the store's own track id hidden in their tags; text search is length-verified and word-order aware; artist-less titles retry inside a BPM window with your library's key breaking ties. Genre, artist, label, release artwork, BPM and key — the wrong edit never gets tagged.
Open-source Essentia models score genre, moods (Party, Dark, Melancholic, Epic, Deep…), vocal components (Vocal / Instrumental, voice gender, instruments), situation and a DJ-style 1–10 energy — from the audio itself. No cloud, no tokens, no uploads.
Energy analysis snapped to rekordbox's own beat grid proposes hot cues — Intro, Build, Drop, Break, Outro — that land exactly on the downbeat. Audition each one on the built-in waveform player before a single cue is written.
Pick a track, a selection or a whole playlist and the similarity engine ranks your library by mood, themes, instruments, BPM (half/double-time aware), Camelot key, genre and energy — each match with a readable reason. Send the results to an existing rekordbox playlist or a new one.
Fix a title, pick a genre, add or remove tags, change a track's energy — right in the library grid. Every edit is staged, marked and revertable, and a tag-confidence setting (strict / balanced / open) re-tags the whole library instantly from the stored scores, no re-analysis.
Files with missing or wrong metadata are identified Shazam-style: an acoustic fingerprint looked up in the open AcoustID database recovers the real artist and title — even for unreleased mixes no store sells. Artwork falls back through four sources, ending at the cover embedded in the file itself and the iTunes catalogue.
Anything the matcher refuses gets a live candidate picker: Beatport's own waveform next to your file's, a two-minute store preview per candidate, lengths, BPM and key with agreements in green. One click writes the match. Tags land in My Tags, in Comments (Ambient, Deep // Warmup), or both.
Standalone mode analyses any folder of music — drag it in and Crate Control works out the BPM, key, beat grid and waveform itself, then runs the same AI passes and cue detection. The result exports as a rekordbox XML you can import. Your files are only referenced, never copied or modified.
Browsing and the full AI analysis are free in both modes, forever. Applying to rekordbox, hot cues and XML export unlock with a 14-day trial — email only, no card — and after that with a one-time licence, yours for good. Nothing you analyzed is ever lost: it all applies the moment a key goes in. Licences work offline for weeks at a time, because gigs don't have wifi.
Requires macOS and rekordbox 6 or 7.