Natural Language Playlist is a web tool that builds music playlists from free-form text descriptions rather than requiring a listener to search by artist or song title. It was built by Abelardo Riojas, a data science graduate student, with a beta version active by early-to-mid 2023.
Users describe the mood, genre, lyrical themes, or cultural vibe they want, for instance a description of a memory or feeling rather than a specific song, and the tool uses sentence embeddings matched against curated song metadata to surface a fitting track list. People can use it as a guest or log in with Spotify to save the generated playlist directly to their account, and login data feeds back into the model to refine future recommendations; the tool works best with positive descriptions rather than negations, writing 'quiet rock songs' instead of 'rock songs that are not loud.'
Natural Language Playlist suits music listeners who want to discover new songs through descriptive, mood-based search rather than people who already know what artist or track they want to hear. The tool is free to use, with no subscription or paid tier, and stands as a solo, independently built project rather than a funded startup.
