Google expands Lyria 3 Pro: its AI composer now generates tracks up to three minutes long

Google has updated its music model and introduced Lyria 3 Pro, a far more capable version that generates tracks up to three minutes long. Until now, Lyria was strictly limited to 30-second clips: enough to demonstrate what it could do, but clearly not enough for real content production. The difference matters. Thirty seconds is a snippet; three minutes is a full composition. A sixfold increase in length moves the tool from demo territory into working-tool territory for content creators.

The main change is not only length but structural control. Lyria 3 Pro lets users explicitly specify which elements a track should contain: intro, verse, chorus, bridge. That changes how the tool works. Instead of describing a mood and waiting to see what comes out, you can build a complete musical structure. For anyone with even basic familiarity with music production, this is a significant improvement over competitors, which mostly leave arrangement to the model with no way to influence it directly.

Otherwise the workflow is standard for the market: you describe the genre, mood and instrumentation, and the model creates a track. Suno and Udio, which first popularized generative music at scale, work the same way.

Lyria 3 Pro can also write song lyrics from a prompt. But it has one unusual feature: creating music from a reference image. You upload a photo or a cover, and the model generates a fitting soundtrack. It is an atypical approach, but potentially useful for videographers, marketers and designers who need music for a specific visual idea.

Lyria's key differentiator against competitors is its built-in integration with the Google ecosystem. Lyria 3 Pro will appear in several of the company's products, although the specific list has not yet been disclosed. Independent services such as Suno or Udio have to fight for distribution on their own, spend on marketing and convince users to move to a new platform. Google is in a different position: it can simply add a "create music" button to YouTube, Google Slides or any other service with an audience in the hundreds of millions. That is its main trump card, which comes down not only to model quality but to the scale of distribution.

Lyria 3 Pro is part of Google's consistent strategy of covering every format of generative media. The company already has Veo for video and Imagen for images. Music AI was arguably the only major media category where Google clearly lagged: while Suno and Udio attracted tens of millions of users, Google stayed on the sidelines. The launch of Lyria 3 Pro with expanded capabilities looks like a serious bid for this segment, not just a technical update.

Three minutes is already a practical working format. Background music for a YouTube video, an advertising jingle, a podcast intro, a soundtrack for a presentation or a corporate video: all of these fit within a three-minute limit. For content creators who currently either pay for licensed music or rely on limited free libraries, a built-in generator inside a familiar service could become a real solution. That is especially true if Google builds Lyria into YouTube Studio or similar tools for creators.

The generative music market is still small but growing fast, and Google enters it with an obvious structural advantage. Suno and Udio have shown that demand for such tools exists, and it is considerable. Now they are joined by the world's largest media platform, with distribution built in. If Lyria 3 Pro turns out to be good enough, it risks becoming the default tool for millions of users, not because it is better than its competitors, but simply because it will be where people already work every day.