Google expanded its Music AI Sandbox with Lyria 2 and broader access for US-based artists. The tools focus on three practical actions. They create new musical ideas, extend existing audio, and edit parts of a track.
Create generates new ideas
The Create tool lets users describe a sound and generate short musical ideas. The source says prompts can specify genres, moods, instruments, and vocal styles, while users can also add lyrics and guide elements such as tempo and key.
Music AI Sandbox originated through YouTube's Music AI Incubator and was developed in collaboration with musicians.
Extend and Edit work from existing audio
Extend lets a creator upload a clip and ask the model to continue the musical idea.
Edit lets users change the style, genre, or mood of a clip, either across the whole piece or in selected parts. Text prompts and preset controls can be used to reshape or transition the material.
Those tools make the Sandbox more useful for musicians who already have source material and want to modify it rather than generate everything from zero.
Access expanded to more US artists
At the time of the announcement, Google said the upgraded Sandbox was available to more artists in the United States and that it was collecting feedback as the tools developed.
The official details are available from Google DeepMind.
