Google’s Gemini Chatbot Adds Music Creation Feature with New Lyria 3 AI Model

Google has added AI music generation to its Gemini app, powered by the new Lyria 3 model. Users can now create short, custom tracks using simple text prompts, with built-in watermarking for transparency.

Google has expanded the capabilities of its Gemini AI chatbot by enabling it to generate custom music tracks, thanks to the integration of its latest generative audio model, Lyria 3. This update brings music creation directly into the Gemini app, allowing users to generate personalised 30-second tracks using simple descriptions or images.

With this feature now in beta, anyone with the Gemini app can type a prompt like “uplifting indie pop for a weekend road trip” or upload a photo or video, and Gemini will compose an original audio piece that includes melody, rhythm, and even lyrics if desired. Generated tracks also come with automatically produced cover art to make sharing easier.

Lyria 3 represents Google DeepMind’s most capable music generation model to date, offering enhanced creative control such as specifying tempo, vocal style, and genre. Google emphasises that this tool is designed to help users explore musical ideas and express themselves, rather than to replace professional music production.

To promote transparency and responsibility, all AI-generated music includes an imperceptible watermark called SynthID, which helps identify content created by the model. Gemini can also check whether an uploaded audio file was generated using Google’s AI.

The rollout is global, available to users aged 18 and older in languages including English, Spanish, French, Hindi, and more. Premium subscribers of Google’s AI plans can generate longer or more frequent music outputs. Additionally, Google is planning to integrate Lyria 3 into features like YouTube’s Dream Track, making it easier for creators to pair custom soundtracks with their video content.