Spotify Adds AI Podcast Search via Text Queries

Spotify has extended the Prompted Playlists feature to podcasts — now Premium users can discover episodes through simple text queries instead of endlessly browsing the catalog manually. The Prompted Playlists feature first appeared in beta in December of last year, but at that time worked exclusively with music. The principle is simple: a user enters a text query — for example, "psychology podcasts for a morning run" or "detective stories for an evening" — and the algorithm creates a personalized playlist, "orienting" itself toward the specified genre or mood.

In essence, it's an analog of Discover Weekly, but manually controlled. The April 7 update applies the same mechanics to podcasts. The algorithm analyzes the query and compiles a selection of specific episodes — not just full shows, but individual episodes that fit the theme.

This fundamentally changes the search logic: instead of first finding a suitable show and then selecting the right episode, the user immediately gets a ready-made list. For now, the feature's capabilities are limited: the beta version is available only in English and only for Premium subscribers in the US and Canada. Spotify has not announced when or in which regions a full launch will occur.

For Spotify, this is a logical step toward strengthening personalization through generative tools. The company has long used machine learning for music recommendations, and now is applying a similar approach to podcasts — a segment where competition with Apple Podcasts and YouTube remains intense. Text prompts lower the barrier to entry for new listeners and potentially increase time spent in the app.