Google (GOOG)(GOOGL) has added a custom music generation feature to its artificial intelligence assistant app Gemini.
The tech giant is rolling the feature out today in beta on the Gemini desktop app. Users have multiple avenues by which to generate music. It can be done through text, image, or video. It then creates 30-second music tracks with custom cover art by its AI image generator Nano Banana. The feature also generates lyrics for the short musical piece if wanted.
The new custom music feature is powered by Lyria 3, which is Google DeepMind’s latest generative music model. It is available today in English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese. Google plans to expand quality and coverage of more languages. It will roll out on the mobile app within the next several days.
Lyria 3 is also available on YouTube’s Dream Track, where creators of “Shorts” videos can use it to add their own custom music.
Even though some music might be inspired by real artists, it is not intended to mimic artists or infringe on existing content. Google even recommends artists or users report any instances of a perceived copyright violation.
Other music platforms have also used AI technology to enhance their products. For example, on Monday, Apple (AAPL) released the beta version of its iOS 26.4 for developers. It can create an AI-generated playlist for Apple Music subscribers based on their music listening history.
This is similar to a feature Spotify (SPOT) added for its Premium subscribers back in 2024.