How Spotivibly Generates Playlists
Spotivibly builds a playlist in three steps: AI research proposes tracks from your prompt, each track is checked against the Spotify catalogue, and a refinement pass replaces the ones that do not resolve. It usually takes under two minutes and does not measure tempo or audio features.
Every Spotivibly playlist runs through a four-stage pipeline: AI research finds candidate tracks, Spotify validation verifies they exist, AI refinement fills any gaps, and export saves the result to your library.
The whole flow takes about 60 seconds. You describe a vibe in plain language; AI handles the catalog work. Spotify-verified, YouTube-compatible, and shareable.
Stage one: research
An AI model proposes candidate tracks from your prompt. Mood, genre and multi-artist prompts are routed to a model with web search grounding, while artist-specific requests use a search-backed model that is better at niche and regional catalogues.
Stage two: validation
Every proposed track is looked up in Spotify's catalogue. This is the stage that removes invented titles and wrong-artist matches, and for artist-specific requests it mines that artist's albums, singles and collaborations directly rather than relying on what the model recalled.
Stage three and four: refinement and save
Anything that failed validation goes back for a refinement pass with the failures named, so replacements are targeted rather than random. What survives is searched on your chosen platform and saved as a playlist, on Spotify or YouTube.