Google launches Lyria 3, an AI music tool in Gemini for creating 30-second tracks from text, images, or videos, with AI-generated lyrics, vocals, and art, available now on desktop and coming to mobile, free with limits, and watermarked for identification.

Google has launched Lyria 3, an AI music generation model integrated into Gemini, enabling users aged 18 and over to create 30-second original tracks from text, images, or videos. Available now on desktop and coming to mobile, the feature generates music with lyrics, vocals, and AI-created artwork using prompts for mood, genre, or scenes. Supported in multiple languages, it includes SynthID watermarks to identify AI origin and is free with usage limits for basic users, higher limits for paid subscribers. The tool is designed for creative expression, not professional use, and avoids direct artist imitation. It will also be added to YouTube Shorts’ Dream Track.