Google puts Lyria 3.5 music model in the Gemini app and API

4 Sep 2026: Google published that Lyria 3.5 — its music generation model — is now available in the Gemini app and the Gemini API, with genre/vocal-or-instrumental controls, new templates, and short or longer tracks. Company blog is the primary. Quality and availability claims stay Google’s.

4 Sep 2026: Google published “Create your best tracks yet with Lyria 3.5 in Gemini.” The Keyword post announces that Lyria 3.5 is now available in the Gemini app and the Gemini API. That dated company blog is the filing event.

Google calls Lyria 3.5 “our best-sounding music generation model” and says it brings more expressive vocals and richer musical arrangements so users can craft tracks with higher fidelity. Those are Google’s product claims — not an independent listening test.

In the Gemini app, Google says users can select or describe a genre, choose vocal or instrumental styles, use new templates (from background music to custom birthday tracks), and choose short or longer tracks.

Availability on the post: available to all users globally on the web and in the mobile app. Google also says it is available for artists and AI creatives in Google Flow Music, and for developers and technologists through Google AI Studio and Google Vids.

This filing does not invent sample rates, watermarking or SynthID details, pricing, or artist-voice bans — those do not appear on this primary. Distinct from the already-filed Gemini Windows desktop app, agentic video understanding, Dreambeans US rollout, and speech-to-text transcription stories.

A dated Google primary moves Lyria 3.5 from Flow Music into the main Gemini app and API surface — music generation as a default Gemini feature for every user, not a niche creative sidecar.