Spotify launches an AI DJ that talks between your favorite tracks

Opening a music app only to stare at millions of tracks without knowing what to play is a universal modern headache. Algorithmic playlists often feel cold, and manual curation gets tedious fast.

Spotify just unveiled DJ, a personalized AI guide currently rolling out in beta. Instead of just handing you an endless queue of tracks, it curates a custom music stream and talks to you between songs with commentary delivered in a remarkably realistic voice.

A three-part recipe for a personal radio station

The feature combines three distinct layers of technology to turn your music stream into something resembling a late-night radio show.

At the core sits Spotify’s traditional personalization engine. It scans brand-new releases it thinks you will enjoy and resurfaces forgotten favorites from years past. If the mood is off, tapping the DJ button at the bottom right immediately pivots to a different genre or vibe.

The spoken commentary comes from generative AI powered by OpenAI. Spotify put these generative tools in the hands of its human music editors, scaling their cultural insights into bite-sized stories about artists, tracks, and genres.

To actually speak those lines, Spotify used the voice platform from its Sonantic acquisition. The debut voice model is based on Xavier "X" Jernigan, Spotify's Head of Cultural Partnerships and a former host of its morning show The Get Up, giving the AI a relaxed, conversational tone.

The feature initially rolled out in English to Spotify Premium subscribers across the U.S. and Canada on iOS and Android. Rollouts expanded to the UK and Ireland in May 2023, hit 46 more markets in August 2023, and later expanded in March 2026 to include English and Spanish options across European countries and Japan.

Why it matters

For everyday listeners, DJ turns passive music streaming into an active, hosted listening experience without requiring any manual effort. You get the context of a radio DJ paired with an algorithm that actually knows your listening history. The barrier, of course, is that it requires a paid Premium subscription and rolled out with limited language support.

For the music industry, this changes the personalization playbook. Competing services like Apple Music have spent years optimizing recommendation algorithms, but those recommendations still arrive as silent, impersonal playlists. By introducing human-style voice commentary, Spotify effectively reinvents terrestrial radio for the streaming era.

We think Spotify's execution here is particularly sharp. Instead of letting an LLM hallucinate music trivia or trying to replace music editors, it used OpenAI's tech as a force multiplier for its editorial staff. Sonantic provides the warmth, while editors supply the expertise. The main risk looks like conversational fatigue: if the DJ talks too frequently or interrupts high-tempo sessions with generic trivia, users will simply tap past it to keep the music playing.