Systems and processes are disclosed for operating a digital assistant for media search and playback. In an exemplary process, an audio input containing a media search request can be received. A primary user intent corresponding to the media search request can be determined and one or more secondary user intents based on one or more previous user intents can be determined. A primary set of media items corresponding to the primary user intent can be displayed and one or more secondary sets of media items corresponding to the one or more secondary user intents can be displayed.