Gemini Can Now Generate Music From Text and Photos

The line between text and sound just got thinner: Google has built a music generation tool into the Gemini app, powered by the Lyria 3 model. Any user can now describe the sound they want in words, or upload a photo, and get a finished 30-second track. No sheet music, no studio knowledge, no special equipment. This is not just another item in a long list of updates. It is an attempt by Google to redefine who gets to call themselves a music author at all.

To grasp the scale of the move, some context helps. Text-to-audio generation has existed for several years: Suno, Udio and Meta's MusicGen all offered similar capabilities at varying levels of quality. But most of these services lived apart from mainstream products, required sign-ups in specialized apps, and remained a niche hobby for technically savvy users. Google is betting on something different: Lyria 3 is built directly into Gemini, an app used by hundreds of millions of people worldwide. The barrier to entry almost disappears.

Lyria 3 is Google's most advanced music model to date. The company developed it within DeepMind's research program, and the result is now moving from the lab into the pocket of an ordinary user. The mechanics are simple: you describe a mood, genre, instruments or atmosphere in text, for example "relaxing lo-fi with piano and rain outside the window", and the model generates a track. The alternative route is even more interesting: you can upload an image, and Lyria 3 interprets its visual content as a musical picture. A sunset over the sea becomes one melody, city bustle becomes a completely different one. This multimodal approach is what sets Google's offering apart from most competitors.

For the industry, this signal has several layers. The first is obvious: the largest tech companies are competing seriously for the creative audience. OpenAI has already integrated image generation into ChatGPT, Meta is developing its own multimodal tools, and Apple is quietly building AI features into its ecosystem.

With Lyria 3 in Gemini, Google makes music the next battlefield. The second layer is more complicated: the arrival of such tools in a mass-market product inevitably raises questions about copyright and monetization. What data was Lyria 3 trained on?

What happens to a track you generate: can you publish it, sell it, use it in commercial projects? So far Google has not officially given exhaustive answers, and this remains an area of uncertainty that the industry will watch closely.

For the average user, the consequences are far more straightforward. A podcaster can create a unique intro in a minute. A video creator can get background music without hunting for Creative Commons-licensed tracks. Someone who has had a melody playing in their head for years can finally bring it to life without knowing a single note. This audience, not professional musicians but millions of people with creative needs and no technical means, is Google's real target. Thirty seconds is admittedly not much, but it is a perfectly adequate format for jingles, intros, atmospheric inserts and experiments.

Looking ahead, the right stance is neither euphoria nor skepticism. Music generation is still far from threatening professional composers, just as text AI has not displaced journalists. But it changes the economics of creativity: it lowers the cost of producing content, widens the circle of people who can create it, and creates new professional roles for those who can write good prompts and edit the results. Lyria 3 in Gemini is not the end of the music profession, but the start of a conversation about what it means to be an author in an era when the instrument can play itself.

*Meta is designated an extremist organization and banned in the Russian Federation.