Google adds music generation to the Gemini app using DeepMind’s Lyria 3

Summary

Google announced it is adding a music-generation feature to the Gemini app, powered by DeepMind’s Lyria 3 music model, which is still in beta.

The feature can generate a 30-second track with lyrics from a text description, and it also produces cover art. Google says you can upload a photo or video and have Gemini create a song that matches the media’s mood.

In practice

You describe the kind of song you want, and Gemini generates a 30-second track along with lyrics and cover art (described as made by “Nano Banana”).

Google says users can change and control elements such as style, vocals, and tempo.

Context

Alongside the Gemini rollout, Google is making Lyria 3 available to YouTube creators via Dream Track and says it is expanding Dream Track availability globally (it had previously been available only to creators in the U.S.).

Google says you can’t outright mimic an artist, but including an artist’s name can be treated as broad inspiration to create a similar style or mood. The company also mentions having filters to check outputs against existing content.

Why it matters

• Music generation inside Gemini adds an end-to-end “prompt-to-track” workflow, including lyrics and imagery.
• SynthID watermarking and in-app detection point toward clearer labeling and verification of AI-generated music.
• Expanding Dream Track beyond the U.S. could widen access for creators and increase the volume of AI-generated tracks.