Available to download while your membership is active · Not sold separately
· macOS, Intel and Apple Silicon
At a glance
Work out what to listen to next.
It is not a music player. What it does is the part that comes before, working out what
to listen to next, then handing you over to Spotify or Tidal or YouTube or wherever you
already listen.
01
Start with something you like
Search a track or artist and see where it sits, with the source of its genre.
02
Follow it outward
Similar artists come from listeners and shared labels, each with its reason.
03
Keep what you find
Save tracks into lists, correct genre tags and export when you want them elsewhere.
See it in action
From a search to a list worth keeping
Real screens from the app. Select any screenshot to enlarge it.
Search results
Step 01
Start with a track or artist
Search results include 30-second previews, iTunes purchase links and a button to
open the track in your streaming service.
Similar artists
Step 02
Follow the thread outward
Similar artists come from shared record labels and from what listeners play
together, each labelled with its source.
Lists
Step 03
Keep what you find
Organise discoveries into lists by genre, with playlist import and track counts.
More screens
Track previewAbout the artistList detailSettingsDark theme
Included with the $49 annual membership · Members only
When you want the specifics
All the useful detail, easy to find.
The complete feature list, how the app treats its data, and answers to common questions.
Open any heading for the full answer.
Everything included
Local-first macOS desktop app for Intel and Apple Silicon
Search by track or artist, with 30-second previews from Apple alongside the
Apple store link
Links out to Spotify, Apple Music, Tidal, Amazon Music and YouTube
Lists you can build by hand or import from a Tidal export, an M3U file, a CSV or
plain text
Genre tagging you can correct, and export to CSV, M3U or plain text
Discography panel with release detail, credits you can follow, and label
exploration
Style primers that explain what a genre sounds like and where it came from
Artist facts from Wikidata, and optional AI notes using your own Anthropic or
OpenAI key
No account, no sync, no telemetry
How the main features work
Start with something you like
Search by track or artist, and the app looks it up across open music databases
and shows what it found, including a 30-second preview where Apple has one.
See where it sits
Genre comes from release data, and you can see which source answered. A
specific style and a broad genre are not the same thing, so a broad one is
never dressed up as precise.
Follow it outward
Similar artists come from two directions: artists that listeners actually play
together, and artists sharing a record label. Each suggestion says which of
the two put it there, so you can judge it.
Keep what you find
Save tracks into lists, tag them by genre, correct any tag the app got wrong,
and export the list when you want it somewhere else.
How it treats the data
Learning about the music, not just finding it
Every artist has an About the artist panel built from Wikidata: where they are
from, when they started, which labels they have released on, who is in the band.
Those are retrieved facts, shown as retrieved, with the source linked. Nothing on
that panel is written by AI.
AI notes are separate and optional, and they describe the music rather than the
person: how it is produced, what to listen for, where the style came from. Every
artist name the model produces is checked against a real music database before you
see it, and where there is not enough to go on the app says so instead of
guessing.
Where it is thin, it says so
Music data is uneven. Well-documented artists have plenty, newer and independent
artists sometimes have almost nothing, and a few have no entry anywhere. Where
that happens, you get told rather than shown a confident blank.
Sources are rate limited too, so working through a long imported playlist takes a
few minutes. There is a progress indicator, and an estimate before you start.
Listening, accounts and AI
Does it play full songs?
No. Apple's 30-second previews are all that plays in the app, and full listening
happens in your own streaming service under your own login.
Do I need a paid streaming subscription?
No. Previews and purchase links work without one, and YouTube is among the
services it links out to. If you do pay for Spotify or Tidal or Amazon Music, set
that as your default and the links go there instead.
Do I need to sign up for anything to use the data?
No. Music data comes from Discogs, ListenBrainz, Wikidata and the iTunes Search
API, and none of them need an account or an API key.
Does it need AI?
No, everything works without a key. AI notes are an extra, using your own
Anthropic or OpenAI key, and the artist facts are not AI at all.
Membership and updates
Is Music Discovery sold separately?
No, it is only available with the membership.
What's changed in the latest version?
Updates are included while your membership is active. You can see the full list of
changes on the
changelog page.
