Google Brings Lyria 3.5 AI Music to Gemini

Google has added Lyria 3.5, its latest AI music generation model, to the Gemini app and Gemini API. The model can create complete songs, instrumentals and soundtracks from text prompts or photos, with richer arrangements, more expressive vocals and better prompt understanding than earlier versions. Users can choose genres, vocal or instrumental output and track length, while templates help beginners make background music, jingles or ringtones. Generated tracks cannot currently be edited mid-prompt. Google says every Lyria song carries a SynthID watermark and the model blocks artist voice cloning and copyrighted lyrics. The launch comes as AI music rapidly expands on streaming platforms, with Digital Trends citing Deezer data showing AI-generated songs account for 44% of daily uploads and noting continuing concerns about fraudulent streams and copyright disputes.

Digital Trends, Manisha Priyadarshini – Google brings its best AI music model Lyria 3.5 to the Gemini app – 5th September 2026