Google has added music creation to Gemini, letting users generate short, custom tracks using text prompts or uploaded images and videos. The feature runs on Google DeepMind’s Lyria 3 model and is now available in beta, offering a new way for users to create and share personalized audio without needing music production experience.
The update expands Gemini’s creative tools beyond image and video generation, giving people the option to turn ideas or visual content into music in seconds.
SEE ALSO: AI music platform SoundBreak launches with licensed artist models
Users can describe a genre, mood, or concept, and the system produces a 30-second track with lyrics or instrumental elements tailored to the request.
Lyria 3 introduces automatic lyric generation, so users don’t need to write words themselves. It also allows more control over musical style, vocals, and tempo, giving creators flexibility when shaping the final result. The model produces tracks with more complex arrangements compared with earlier versions.
Creating music in Gemini
Music creation works in two main ways. Text prompts can be used to generate tracks based on memories, jokes, or themes, while uploaded photos or videos can guide the music’s tone and lyrical direction. The app interprets the content and generates a matching soundtrack meant for quick sharing rather than polished studio output.
Each track includes custom cover art generated by Nano Banana, and users can download the result or share it through a link. The focus is on short-form creativity, giving users a fast way to add music to everyday moments or social sharing without needing advanced editing tools.
Lyria 3 is also being integrated into YouTube’s Dream Track feature. The system supports creators making Shorts by generating custom background audio or vocal sections that match the tone of short videos. The rollout began in the U.S. and is expanding to creators in additional countries.
Google says all audio generated in the Gemini app includes SynthID, an invisible watermark used to identify AI-created content. The app now supports audio verification alongside image and video checks, allowing users to upload files and ask whether they were produced with Google’s AI tools.
The company says it has worked with the music community during development and has tried to address copyright concerns while training Lyria 3. Tracks are intended for original expression rather than recreating specific artists. If prompts mention artists by name, the system interprets them as broad stylistic inspiration rather than direct imitation.
Filters are in place to compare outputs against existing content, and users can report material that could infringe on rights. Access to the tool also requires users to follow platform rules covering intellectual property and privacy.
Lyria 3 music generation is available in the Gemini app for users aged 18 and over across multiple languages, including English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese. It is rolling out first on desktop, with mobile availability expanding shortly after, while paid Google AI subscription tiers receive higher usage limits. Try it out today at gemini.google.com.
What do you think about Gemini’s new music creation feature? Let us know in the comments.
