New model brings longer tracks and deeper control as music AI competition intensifies
Google is expanding its push into AI-generated music with the release of Lyria 3 Pro, a more advanced version of its music generation model that arrives only a month after the launch of Lyria 3. The new model significantly extends the length of the compositions users can create, increasing the maximum from 30 seconds to three minutes, and signals that Google wants to move beyond novelty clips into something closer to structured, usable music creation.
The upgrade matters because duration changes the product category. A 30-second track can work as a quick experiment, a social media snippet or a lightweight creative tool. A three-minute composition starts to look more like an actual song or soundtrack, something that can be used in videos, presentations, demos or more ambitious creative projects. That makes Lyria 3 Pro less of a toy and more of a platform feature with broader commercial value.
Google is also presenting the new model as more controllable, not just longer. Users can now prompt for specific parts of a song, including intros, verses, choruses and bridges, which suggests the model has a stronger grasp of musical structure than its predecessor. In practical terms, that gives creators more direction over how a track unfolds instead of relying mostly on broad stylistic prompts.
Google is building music generation into more of its ecosystem
The launch of Lyria 3 Pro shows that Google no longer sees music generation as a niche experiment inside one app. The model is rolling out inside Gemini, though access there will be limited to paid subscribers, and it is also being integrated into Google Vids and ProducerAI. That wider deployment points to a deliberate strategy: generative music is becoming one more capability that Google wants available across consumer, creator and enterprise products.
The enterprise side is especially notable. Google is adding Lyria 3 Pro to Vertex AI in public preview, as well as to the Gemini API and AI Studio, which means developers and businesses can begin using the model in larger-scale workflows. That moves the product beyond casual consumer experimentation and into areas such as content production, software development and brand-facing creative tools.
In other words, Google is not treating Lyria 3 Pro as an isolated feature release. It is turning it into infrastructure, something that can support both individual creators and professional environments inside the company’s broader AI stack.
The company is trying to balance creativity with safety and IP concerns
As with every major generative media product, Google also has to address the question of training data, imitation and misuse. The company says Lyria 3 Pro was trained using partner data along with permissible material from YouTube and Google. It also says the system does not imitate specific artists, though it acknowledges that prompts referencing an artist can lead the model to take broad inspiration from that style.
That distinction is important but also delicate. Music generation tools are increasingly being judged not only on output quality, but on how clearly they separate influence from imitation. The stronger and more convincing these models become, the more pressure they face from musicians, labels and platforms worried about copyright, attribution and the dilution of artistic identity.
Google has tried to address transparency by marking all tracks generated through Lyria 3 and Lyria 3 Pro with SynthID, a signal that the music was created with AI. That move reflects a wider industry trend as platforms and streaming services look for ways to distinguish machine-generated content from human-made work.
The release lands in a music industry already on alert
The timing of the launch is notable because the music business is becoming more defensive about AI-generated audio. Streaming services and tech companies are rapidly rolling out tools to identify synthetic music, detect misattribution and limit the spread of low-quality AI uploads. That means Google is expanding Lyria just as the rest of the ecosystem is becoming more serious about policing what these tools produce and how those outputs are labeled.
This makes Lyria 3 Pro part of a larger competitive and regulatory moment. On one side, the technology is improving quickly, with better structure, longer duration and more direct user control. On the other, the surrounding industry is trying to build safeguards before the flood of AI-generated music overwhelms discovery, metadata integrity and artist trust.
For Google, the opportunity is clear. If it can make music generation feel useful, controllable and responsibly deployed, Lyria 3 Pro could become one of the more practical creative tools in its AI portfolio. But the challenge is just as obvious. The better these systems get, the less room there is for vague answers about originality, ownership and the boundaries between inspiration and imitation. That tension is likely to define whether AI music becomes a durable category or a legally and culturally contested one.
