Google Brings Lyria 3.5 Music Generation to the Gemini App
Google brought Lyria 3.5, DeepMind’s latest music generation model, into the Gemini app, describing it as its best-sounding music model yet.
Details
- Quality improvements: Lyria 3.5 delivers more expressive vocals and richer musical arrangements, producing higher-fidelity tracks than previous versions
- How it works in Gemini: users select or describe a genre, choose vocal or instrumental style, and can use templates for common use cases like background music or personalized audio, with options for shorter or extended track lengths
- Availability across Google’s products: live now in the web and mobile Gemini app, Google Flow Music for artists, Google AI Studio for developers, and Google Vids within Workspace
- Global rollout: accessible globally rather than limited to specific markets
What happened next
Bringing Lyria 3.5 across consumer, creator, developer, and Workspace surfaces simultaneously signals Google is treating AI music generation as a cross-product capability rather than a single standalone app, competing more directly with dedicated music-generation services.