Veteran internet radio company Pandora needed an inexplicably long time to join the likes of Apple and Spotify with its own premium music streaming platform, but even though the market duopoly seems harder and harder to break up, efforts aimed at just that are certainly ramping up.

In addition to supporting devices as diverse as Amazon’s Echo lineup, Google Home smart speakers, Google Assistant-powered smart displays, and the Apple Watch, the latter of which recently received offline listening capabilities, Pandora Premium also features a native new Voice Mode for differentiation in this increasingly crowded industry.

Basically, Pandora built its very own Siri, Alexa, and Google Assistant-rivaling AI from the ground up, although the latest entry into the smart assistant market can basically only do one job.

Interestingly, while Pandora’s long-term goal is undoubtedly to get more people to pay for its $9.99 a month on-demand service tier, Voice Mode will be made available for both free and premium users on iOS and Android.

For the time being, only “select” listeners will be granted beta access to the intuitive, hands-free music control and discovery feature, with general availability for all mobile users coming “soon.”

The way Voice Mode works sounds pretty straightforward, waking up to the sound of “Hey Pandora” and helping you change stations, control volume, skip or pause music, as well as answering “thematic requests” without you having to lift a finger.

Obviously, the key selling point is not how Pandora’s voice assistant handles basic navigation commands, but rather how it can “continually refine” your listening experience by detecting and understanding natural speech.

Not only can you ask your new digital assistant to play a specific song, station, podcast, or playlist, but also rely on it to deliver personalized music based on “each user’s unique tastes, moods and favorite activities”, as well as handle open-ended, interactive, and directional requests.

Pandora claims Voice Mode will be able to serve results tailor-made to each individual user to general requests for new music or similar tunes to the ones already playing. In other words, the app will try to guess what you might like to hear based on your listening history, using “predictive understanding of individual listener preferences.” Sounds pretty cool, but we’ll have to wait and see how good of a job it can do in real-life scenarios.