CONFIRMED: Google brings Lyria 3.5 to Gemini and the API

Google announced on September 4, 2026 that Lyria 3.5, its music generation model, has arrived in the Gemini app and the Gemini API. The news was reported by Android Authority, in a piece titled "Gemini hands Lyria 3.5 the mic as the music model makes its debut on the app", and confirmed by Unite.AI, TheNextWeb and PPC Land. Worth being precise here: Lyria 3.5 did not debut in September, it first launched on July 29, 2026 inside Google Flow Music, a separate and more limited product. What happened now is a platform expansion, the same music model moves out of an isolated app and becomes available across the broader Gemini ecosystem, developers included.

Who gets access, and where

Starting September 4, Lyria 3.5 music generation is available to all Gemini users worldwide, both on web and on the mobile app. For developers, access goes beyond the consumer app, Google also opened Lyria 3.5 through the Gemini API, Google AI Studio and Google Vids. In other words, the same capability a regular user triggers by asking Gemini for a song can be called programmatically by any technical team inside its own product flow.

What the model delivers: up to three minutes of cohesive music

Google describes Lyria 3.5 as its best sounding music model yet, with more expressive vocals and richer musical arrangements, plus higher fidelity than earlier Lyria versions. In practice, the most concrete upgrade for anyone using the tool is duration: Lyria 3.5 supports variable song length, from a short 60 second clip up to a full three minute song, generating a cohesive track that matches the requested duration instead of simply looping a snippet. The Gemini app also gained new controls, users can pick or describe a genre, choose vocal or instrumental only, set short or long tracks, and start from ready made song templates.

Why it matters for Brazilian agencies and SMBs

For anyone producing marketing content in Brazil, having a track of up to three minutes generated directly through the Gemini API changes the math on soundtrack production. An agency can generate jingles, background tracks for Reels, TikTok and video ads, and branded audio content without relying on paid stock music libraries, and can slot that generation into the same content pipeline it already uses for other Gemini text and image tools, automating production end to end. That is a real gain in speed and cost for anyone delivering high volumes of short video pieces every day.

The warning that cannot be skipped: commercial usage rights

Caution is warranted here. Google did not publish, alongside the September 4 announcement, a detailed commercial license spelling out how Lyria 3.5 generated audio may be used in paid campaigns. The model applies safety filters that block requests to imitate specific singers' voices or reproduce copyrighted lyrics, and Google states it trains Lyria on licensed data, but neither of those amounts to an explicit public statement that every generated track can be freely used in commercial ads. Before putting any Lyria 3.5 track into a paid client campaign, the safe move is to read the current terms of use for the Gemini API and Google AI Studio at the moment of generation, since a flow designed around free personal use can carry different conditions than a redistributed commercial use.

The blunt read

Google is not just testing a toy inside a side app, it is putting professional grade music generation into the same API that already delivers text, image and video, signaling a clear intent to make Gemini the single content production hub for businesses. For marketing and automation teams, the opportunity is immediate, the risk is treating AI generated audio as automatically free of commercial restriction before confirming that in writing in Google's own terms.

Sources

Android Authority, Gemini hands Lyria 3.5 the mic as the music model makes its debut on the app: https://www.androidauthority.com/latest-lyria-model-comes-to-gemini-app-3707842/ | Unite.AI, Google Brings Lyria 3.5 Music Generation to the Gemini App and API: https://www.unite.ai/google-brings-lyria-3-5-music-generation-to-the-gemini-app-and-api/ | PPC Land, Google's Lyria 3.5 puts full-length AI songs in Gemini and the API: https://ppc.land/googles-lyria-3-5-puts-full-length-ai-songs-in-gemini-and-the-api/