Spotify Radio runs on co-listen data. We run on sonic DNA.
Spotify Radio's recommendations come from collaborative filtering. It watches what hundreds of millions of listeners play in sequence and shows you tracks that other people-like-you played after the one you just heard. That works beautifully for mainstream artists with deep co-listen graphs. It struggles when the song is unusual, when the connection is sonic rather than social, or when you want to bridge between genres that don't share an audience.
We don't look at what other people played. We use a language model that analyses the actual sound of your seed — production techniques, tempo, instrumentation, vocal treatment, the emotional register a fan would point to as the song's signature — and finds tracks that share those qualities. Even from artists who have nothing in common in Spotify's graph.
Different inputs, different blind spots. Spotify Radio is excellent at more songs the same crowd likes. We're excellent at more songs that feel the same.
We tell you why each track fits.
Every match comes with a sentence explaining the connection: the production choice, the BPM range, the emotional arc, the specific thing that makes it kindred. Algorithms can't articulate that — collaborative filtering doesn't know why two tracks belong together, only that they co-occur. The "why" is the part most users tell us they screenshot.
You're not locked to one platform.
Spotify Radio lives inside Spotify. You can paste a Spotify, YouTube, SoundCloud, Bandcamp, Audius, or Audiomack link into Match This Vibe — or just type a song name — and every match comes with one-click links back to Spotify or YouTube. The vibe doesn't care which app you opened first.
Cross-genre bridges that algorithms refuse to make.
"Find me ambient that hits like this techno track." "Jazz with this rap song's energy." Spotify Radio refuses — there's no co-listen data crossing those boundaries, no neighbourhood in the graph to recommend from. The model lives for this exact kind of question. It's also the easiest way to fall in love with a genre you didn't know you liked.
The hidden-gems dial Spotify hides.
Their algorithm gravitates toward safe popularity by design. Keep listeners on familiar artists, keep ad-supported tracks in rotation, keep the same recognisable names on the playlist. We give you a slider: Well-known on one end, Hidden gems on the other. Drag it toward hidden gems and the recommendations explicitly skip the obvious picks — you get the emerging artists who genuinely fit but haven't built enough listener-graph data to surface in Spotify's recommendations yet.
Multi-seed intersection.
Spotify Radio is single-seed by design. You give it one track and it generates an infinite stream from that seed alone. We let you paste 2–5 tracks and find the shared centre — the music at the intersection of everything you gave us. That's often a far more precise read on your taste than any one seed track can express alone.