Gemini has been rapidly improving across text, image, video and coding – and now it’s making a big push into music. With its latest update, powered by Google DeepMind’s new Lyria 3 model, Gemini can generate full songs from a simple text prompt or even an image, including lyrics and layered instrumentals.
Users can create 30-second, high-quality tracks, adjust elements like tempo and style and even use existing music as inspiration. The feature is rolling out in beta across multiple languages via Gemini apps, the web, and YouTube Shorts’ Dream Track tool, and is available to both free and paid users (with credit limits depending on the plan).
Google says it has taken steps to avoid copyright issues – tracks include a hidden SynthID watermark and are checked to ensure they don’t mimic real artists or existing songs.
Source: Android Authority
