Google Launches Lyria 3.5 Music Generation in Gemini App for Small-Business Audio

On September 4 2026, Google announced that its Lyria 3.5 music‑generation model can now be accessed inside the Gemini app and through the Gemini API, extending the reach of the earlier Google Flow Music tool that debuted in July.

Lyria 3.5, a generative AI model from DeepMind, can produce full‑length songs up to three minutes long. The company says the new version delivers higher‑quality sound, richer lyrics, improved musicality, and better vocal performance compared with its predecessor, Lyria 3. The Gemini app accepts text or image prompts and returns instrumentals, vocals, and lyrics that match the requested style.

The interface is designed for non‑musicians. Users can pick a genre, tweak style parameters, and choose between instrumental or vocal tracks. A library of templates helps users start projects quickly, while variable track lengths allow a short jingle or a longer background score to be generated from a single prompt.

"Whether you need a custom backing track for a video, a unique brand jingle, or just a personalized ringtone, it’s easier than ever to bring your idea to life," the release states.

Small‑business owners are the primary target. By generating royalty‑free music on demand, the tool can reduce the need for expensive licensing or hiring a composer. A retail store could create a brand‑specific track for in‑store play, while a service provider might generate a greeting tone for phone calls or a promo clip for social media.

Lyria 3.5 is available worldwide on web and mobile. The Gemini app integrates with Google Flow Music—allowing remixing and sharing of studio‑quality songs—and with Google AI Studio, an IDE that gives developers access to Gemini models. These integrations provide a single workflow for generating, editing, and publishing audio.

The announcement cautions that businesses must mind copyright. If a company intends to monetize content produced with Lyria 3.5, it must ensure compliance with licensing agreements. The model’s output is intended to be royalty‑free for personal and commercial use, but users are advised to review the terms of use for each generated track.

While the Gemini interface is user‑friendly, the learning curve can still be significant. Users unfamiliar with digital audio production may need time to understand how to refine prompts and adjust parameters to achieve the desired sound.

Industry observers see Lyria 3.5 as part of a broader trend toward AI‑driven audio tools that lower barriers to high‑quality production. Platforms such as Suno and Soundraw have already entered the market, but Lyria’s integration with Google’s ecosystem—Gemini, Flow Music and AI Studio—offers a unified experience for creators and developers.

The Gemini API preview for Lyria 3.5 is offered at the same pricing tier as Lyria 3, giving developers access to the same capabilities in their own applications. The API supports text and image prompts and returns a full audio track that can be downloaded or streamed.

In summary, Google’s Lyria 3.5 release brings AI music generation into a consumer‑friendly app and developer platform. The model’s ability to produce full‑length, genre‑specific tracks with adjustable style and length makes it a practical tool for small businesses looking to add custom audio to marketing videos, jingles, in‑store music and more. With global web and mobile access and integration with Flow Music and AI Studio, Lyria 3.5 positions itself as a significant addition to the growing suite of AI audio tools.