Gemini Now Generates Custom Music With Lyria 3

Google DeepMind is moving into the music business. The company rolled out Lyria 3, its latest generative music model, in beta within the Gemini app—allowing anyone to create 30-second tracks using just text prompts or uploaded photos.

"Just describe an idea or upload a photo, like 'a comical R&B slow jam about a sock finding their match' and in a matter of seconds, Gemini will translate it into a high-quality, catchy track," DeepMind wrote in the announcement.

Lyria 3 improves on previous versions in three ways: it generates lyrics automatically based on prompts (no need to write your own), gives users more creative control over style, vocals and tempo, and produces more musically complex tracks. The Gemini app pairs the audio with custom cover art generated by Nano Banana.

The feature is rolling out to users 18+ in eight languages: English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese. Google AI Plus, Pro and Ultra subscribers get higher usage limits.

On the responsible AI front, all generated tracks are embedded with SynthID, DeepMind's watermarking technology for identifying AI-generated content. The company noted that Lyria 3 is "designed for original expression, not for mimicking existing artists"—if a prompt names a specific artist, the model takes it as "broad creative inspiration" rather than generating in that artist's style.

Lyria 3 is also available on YouTube's Dream Track, expanding to more countries after launching in the US.