Google Gemini can now generate AI music tracks
Google just rolled out Lyria 3, its latest AI music tool, inside the Gemini app.
Now in beta, it lets you whip up 30-second tracks just by describing what you want or uploading a photo or video.
It even auto-generates lyrics and cover art with a quirky tool called Nano Banana.
You can check if a track is AI-generated
With Lyria 3, you can tweak the style, vocals, and tempo to match your vibe.
It's open to anyone 18+ worldwide and supports eight languages—including Hindi and Japanese.
Every track gets a SynthID watermark so people know it's AI-made, plus you can upload a file and ask Gemini if it was generated using Google AI; Gemini will check for a SynthID and use its reasoning to return a response.
Competing with startups like Suno and Udio
Gemini is now the first big-name AI assistant to build music generation right in—putting it head-to-head with startups like Suno and Udio.
The short 30-second format is strategic, long enough for social media clips and short enough to reduce some copyright risk.
Bonus: YouTube's Dream Track feature (powered by Lyria) is rolling out to creators in other countries.