Google’s AI music maker is coming to the Gemini app

Google has integrated DeepMind's latest audio model, Lyria 3, into the Gemini app, enabling users to generate 30-second AI music tracks based on text prompts, images, or videos without leaving the chatbot. This feature is rolling out globally starting today in eight languages, restricted to users aged 18 and over. Users can request songs based on genres, moods, or memories, with the AI capable of creating instrumental audio or songs with automatically composed lyrics. Google emphasizes that the goal is fun self-expression, not creating masterpieces, and the tracks will include custom cover art generated by Nano Banana for easy sharing. While Lyria 3 was previously limited to Google Cloud's Vertex platform, its expansion to Gemini makes it more accessible, though it follows competitors like TikTok and Microsoft Copilot. Google also noted it is being mindful of copyright, ensuring the tool focuses on original expression rather than mimicking specific existing artists, although it can generate music in a similar style.

(Source：The Verge)