Ever wish your life had background music like a movie? Now it kind of can. Google has added a new feature to its Gemini AI that lets you create your own songs just by typing an idea. Powered by Google DeepMind’s latest model, Lyria 3, this new tool can turn simple prompts into actual music tracks, complete with lyrics and even cover art. What is Lyria 3, and why does it matter Google’s new model, Lyria 3, is an upgrade over its earlier music AI tools. The big difference this time is how easy and realistic it feels. Instead of needing technical knowledge, you can just describe what you want. Something like “a chill lo-fi track for a rainy evening” or even “a funny love song about socks”—and the AI will generate a 30-second track for you. It doesn’t stop at just music. The system can also: Google says the focus is not professional production, but fun, personal creativity.
How to try it for free Lyria 3 is now available inside the Google Gemini app. Right now, it’s rolling out on desktop first, with mobile support coming soon. Anyone aged 18 or above can use it for free. To access it:
The feature currently supports multiple languages, including Hindi, English, Spanish, French, and more. Paid subscribers (AI Plus, Pro, Ultra) will get higher usage limits, but the core feature remains free.
Also read: All AI-created audio, video, images must be labelled starting today, removal of deepfakes in 3 hours
How it actually works Lyria 3 works like a creative assistant. You give it a prompt, and it builds a song around it. You can type a text idea, upload a photo/video, or combine both. For example, if you upload a travel clip, the AI can generate a matching soundtrack in seconds. This makes it especially useful for social media creators who want quick, custom background music. There’s also a template feature if you’re not sure where to start; it suggests ideas and styles automatically. Not just Gemini, but also coming to YouTube Google is also expanding this feature beyond Gemini. Lyria 3 will power the YouTube Dream Track tool, which helps creators generate music for their videos. Earlier limited to the US, this feature is now being rolled out globally. This could make it much easier for creators to avoid copyright issues while still getting custom soundtracks.
Built-in AI watermark for transparency Every track created using Lyria 3 will include a SynthID watermark. This helps identify that the content is AI-generated. Users can even upload a track to Gemini and ask whether it was made with AI. What about copyright concerns? AI-generated music is already a controversial topic. Many artists and music companies have raised concerns about how these models are trained. Some believe AI tools use copyrighted music without proper permission. Google addressed this by saying: We’ve been very mindful of copyright and partner agreements as we’ve trained Lyria 3. The company also clarified that the tool is not meant to copy specific artists. If you mention an artist in your prompt, it will only take general inspiration, not replicate their voice or style exactly. At the same time, musicians can report content if they feel their rights are being violated. Also read: Was Epstein working on memory-erasing technology? What is ‘trans magnetic stimulation’?
All-in-one creative platforms With Lyria 3, Google is pushing AI further into creative spaces. From text and images to music, tools like Gemini are slowly becoming all-in-one creative platforms.
