Google DeepMind has unveiled its AI music generation model, Lyria 3, which has been integrated into the Gemini chatbot.
This neural network can generate audio using text prompts, images, or videos.
"Simply describe your idea or upload an image, like: 'a comical slow R&B track about a sock that found its pair,' and within seconds, Gemini will transform it into a high-quality, memorable composition," the announcement states.
Compared to previous versions, Lyria 3 has been enhanced in three key areas:
- No need to write your own lyrics—the LLM will create them based on the prompt;
- Creative control over style, vocals, and tempo;
- The ability to produce realistic and musically complex tracks.
Gemini generates 30-second audio clips with a custom cover created by Nano Banana, which can be shared with friends.
"The goal is not to create a musical masterpiece but to provide a fun and unique way of self-expression," Google noted.
The model is available in English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese. The initial launch was on desktop, with a mobile version set to be released in the coming days. Subscribers to Google AI Plus, Pro, and Ultra will receive expanded limits.
AI Labeling
AI-generated music is becoming increasingly common. The streaming service Deezer boasts 9.7 million paid subscribers and has reported over 50,000 AI tracks being uploaded daily—about a third of its total uploads.
Moreover, 97% of listeners cannot distinguish AI-generated songs from those written by humans.
All tracks from Lyria 3 come with an embedded label, SynthID—a subtle watermark that allows for the identification of AI content.
As a reminder, in September 2025, Suno introduced the "world's first" generative DAW—Suno Studio, which "radically rethinks the music creation process."