Google Releases Lyria 3.5 Music Generation Model
Google has released Lyria 3.5, the latest version of its music generation model. The new release introduces “Selective Section Painting,” which lets users rewrite specific audio segments without regenerating the entire composition from scratch. It also adds independent control over rhythm and duration for vocals, bass, drums, and other instrument tracks, extends track length to three minutes, and improves vocal synthesis. Google has not disclosed the training dataset composition for Lyria 3.5 — the previous version was trained on licensed YouTube content and partner materials. The model is available on Google Flow Music.