Spotify has begun rolling out a new feature that allows listeners to describe, in their own words, exactly what they want to hear. Known as Prompted Playlists, the update marks a shift in how the streaming service’s recommendations work, giving users a more direct way to shape the algorithm that powers their listening experience.
Initially launching as a beta for Spotify Premium users in New Zealand, the feature builds on the company’s existing AI Playlist tools and is expected to expand to other markets in the future.
How the Spotify music prompt works
Prompted Playlists allows users to type detailed instructions into the Spotify app to generate a personalised playlist. Instead of relying solely on passive listening data, users can actively guide the system with requests such as music from their favourite artists over the past five years or high energy tracks for a specific workout followed by calmer songs for recovery.
Unlike earlier AI playlist tools, the new Spotify music prompt supports longer and more detailed descriptions. Users can also refine the playlist over time, adjusting the prompt to request changes such as fewer upbeat songs or more deep cuts they may have missed.
Listeners can choose how often the playlist refreshes, selecting daily or weekly updates, in a similar way to Spotify’s popular Discover Weekly feature.
How Spotify decides what music fits
Behind the scenes, Spotify draws on a vast amount of listening data to determine which tracks belong in each playlist. This includes a user’s full listening history from their first day on the platform, not just recent activity.
Spotify categorises music using a combination of audio analysis, listener behaviour and contextual signals. Songs are grouped by factors such as tempo, energy, mood, genre and popularity, as well as how often they appear together in user playlists. This helps the system understand not just what a song is, but how and when people like to listen to it.
The new feature also uses what Spotify calls world knowledge. This allows the system to understand references to activities, films, TV shows, emotions or time periods mentioned in a prompt.
The technology powering the experience
The Spotify music prompt is powered by artificial intelligence and large language models, similar to those used in modern chat based AI tools. These models interpret natural language requests and translate them into signals the recommendation engine can use.
Spotify then combines this with its long established personalisation technology, which has been refined over years of analysing billions of daily listening events. Each recommendation is accompanied by short explanations, helping users understand why certain songs were selected.
What comes next for Spotify
Spotify says Prompted Playlists are still in beta and will continue to evolve. Future updates are expected to improve accuracy, expand language support and roll the feature out to more regions.
The move also reflects a wider trend among digital platforms to give users greater transparency and control over algorithms. By allowing listeners to directly influence recommendations, Spotify hopes to make music discovery feel more intentional and creative.
For now, the Spotify music prompt represents a step towards a more collaborative relationship between listeners and the technology shaping what they hear.
