Spotify is expanding the controls around its AI DJ feature with two changes aimed at making the listening experience feel less fixed: selectable voice personalities and a commentary-off mode. The additions are rolling out to Premium subscribers in every region where DJ is already available.

For listeners who enjoy a radio-style stream but do not necessarily want the same synthetic host style each time, the new voice selection is the headline change. Spotify is introducing seven English-language personas and two Spanish-language personas, each with its own name and manner of speaking. The stated range runs from laid-back to gregarious, framing the update as a choice of presentation rather than a change to the music-selection system itself.

The English roster includes three U.S. voices alongside one voice each described as UK, Irish, Australian and Indian. Spanish-speaking listeners get two Mexican voices. It is a meaningful expansion of the DJ’s surface-level identity, even if the available information does not indicate changes to what music the feature can choose, how it responds to requests, or which regions receive individual voices.

What Spotify’s AI DJ is designed to do

DJ began rolling out in 2023 as a personalized, lean-back listening option. “Lean-back” is useful shorthand here: rather than asking a listener to carefully build a queue, browse an album, or repeatedly skip through playlists, the service assembles a run of music tailored to their tastes. An AI voice then appears between selections with context or conversational chatter about the music.

That setup places DJ somewhere between a personalized playlist and traditional broadcast radio. A playlist can be highly specific but silent; a conventional DJ can give a station personality but cannot reshape a program around one person’s listening habits. Spotify’s feature tries to combine the two. It selects music for an individual listener, then adds an audio host intended to make that selection feel like a guided session rather than an unattended stream.

Before this update, the feature had already grown to support requests and additional languages. Requests matter because they make the experience more interactive: listeners can direct the feature to alter the musical direction instead of abandoning the session and manually finding something else. The new update does not replace that behavior. Instead, it gives listeners a choice over how the DJ addresses them—or whether it speaks much at all.

Nine personas, but not necessarily nine different music curators

It is important to separate the voice from the curation. A persona in this context means the character and delivery of the AI voice: its energy, conversational tone and apparent regional identity. It does not inherently mean that each voice has a separate catalog, different recommendation logic, or its own uniquely programmed taste.

Spotify’s description centers on personality, with voices spanning more relaxed and more outgoing styles. That can still make a practical difference. An upbeat, highly chatty delivery can be welcome during a commute or while getting ready for a night out, while a calmer voice may fit focused work, reading, or any setting where interruptions feel more noticeable. The update lets a listener tune that social texture without leaving the DJ product entirely.

The language and regional spread also shows that voice is part of the product experience, not merely an interchangeable button. English users can choose from a wider set of regional options, while Spanish users have two Mexican voice choices. The rollout details do not establish that every listener will interpret every option in the same way, nor do they promise further languages or accents. What is clear is that the feature is moving beyond a single, default on-air identity.

That distinction is worth keeping in mind for anyone expecting a new persona to deliver radically different recommendations. The announced change concerns the narration layer. The personalized DJ mix remains the core product; the voices alter the host who frames it.

The quiet mode may be the bigger quality-of-life change

Spotify is also adding a setting to turn off DJ commentary. With that option enabled, the AI voice will not chat between tracks. It will still speak when the session begins and when the listener specifically directs it to change the music choices.

In other words, this is not a complete removal of the DJ’s voice. It is closer to a reduced-interruption mode. The feature keeps enough speech to identify the beginning of the experience and to acknowledge a requested shift, but it drops the recurring between-song narration that defines the ordinary DJ format.

That difference should be especially useful in group or background-listening situations. Spotify points to dinner-party hosting as an example, and it is an intuitive one: people may want a personalized flow of music without an artificial host periodically cutting into conversations. The same logic can apply to work, studying, cooking, gaming, or any situation where the music should occupy the room without demanding attention.

It also resolves a basic tension in voice-led music products. Commentary gives a service character and can make recommendations more legible, but it can become the least welcome element when users simply want continuous sound. Rather than making listeners choose between a fully narrated DJ session and an entirely different playback feature, Spotify is making the narration adjustable inside DJ itself.

Why customization matters for an AI host

Personalization in streaming is usually associated with the music: recommendations, mixes, discovery shelves and playback queues. DJ adds another category of personalization—the presentation of those recommendations. Voice selection and commentary control recognize that two listeners can want similar songs while wanting very different levels of interaction around them.

That is especially relevant for an AI-generated voiceover. A human radio host has a specific voice and style as part of a station’s identity. A synthetic host is more flexible, but that flexibility only becomes useful when users can exercise some control over it. The new personas turn the voice from a fixed component of the feature into a preference setting, while the commentary control permits an even more restrained approach.

There is also a useful practical split between active and passive listening. An active listener might issue requests, welcome the transitions and treat DJ as a companionable way to explore familiar and unfamiliar music. A passive listener may primarily want the system’s music choices and none of the narration. Neither approach requires a different subscription tier within the announced rollout; both additions are for Premium subscribers in markets where DJ is available.

The change does not answer every question a listener might have. The announcement does not specify where the voice-personality picker sits in the app, whether voices can be changed during a session, or which persona names correspond to each delivery style. It also does not say whether the commentary setting applies permanently after it is selected or must be enabled per session. Those are implementation details users will need to find within the app as the rollout reaches them.

What Premium subscribers should expect

  • Eligibility: The new controls are rolling out for Spotify Premium subscribers.
  • Availability: They are arriving in regions where the DJ feature is already available, rather than constituting a new DJ territory launch.
  • Voice choices: There are seven English personas and two Spanish personas.
  • English lineup: Spotify describes three U.S. voices plus UK, Irish, Australian and Indian options.
  • Spanish lineup: There are two Mexican voices.
  • Commentary setting: Turning commentary off suppresses ordinary between-song chatter, while retaining speech at the start of a listening session and after a user asks for a musical change.

For existing DJ users, the most sensible way to think about the update is as control over ambience. Choose a persona when a hosted music experience sounds appealing; reduce the voiceovers when the recommendations are useful but the room needs fewer interruptions. The music service is not presenting these controls as a wholly new listening product, but as refinements to a feature built around individualized radio-like playback.

Spotify’s broader AI DJ development has already included request-taking and language support. These additions extend that same pattern: the DJ is becoming less of a single preset and more of a configurable listening mode. For more on the wider question of AI organizing user-facing content, see how AI-organized product guidance is being positioned in online communities.

For now, the headline is straightforward. Premium subscribers in supported DJ markets can select from more host voices—or ask the host to largely step aside. That may sound modest, but it gets at the central trade-off of AI-assisted listening: personalized curation is most useful when the listener can decide how present the assistant should feel.