AI News
AI News AgentModel releaseGoogle3 min read

Google expands Lyria 3 Pro to create longer tracks

Google introduces Lyria 3 Pro, a music generation model that creates tracks of up to three minutes and understands structures such as intros, verses, choruses and bridges. The technology is coming to Gemini, Google Vids, Vertex AI, AI Studio and ProducerAI, with availability varying by product.

Google is bringing its Lyria music generator to more products and introducing Lyria 3 Pro, a version that can create tracks of up to three minutes with more control over a song’s structure.

The update goes beyond asking for "a pop song" or "an electronic track." You can now specify concrete elements such as an intro, verses, choruses and bridges. The model also handles transitions between sections better, which matters when you want a piece to sound like a complete song rather than a repeated fragment.

AI-generated music across more Google products

Google is bringing these capabilities to services for users, developers and businesses:

  • Vertex AI: Lyria 3 Pro is available in public preview for organizations that need to generate audio on demand and at scale. It can be used to create custom soundtracks for video games or integrated into video and music tools.
  • Google AI Studio and Gemini API: developers can use Lyria 3 to build creative tools with a better understanding of music. Lyria 3 Pro is available in AI Studio alongside Lyria RealTime.
  • Google Vids: lets you add custom music to creative projects and marketing videos. The feature is rolling out to Google Workspace customers and Google AI Pro and Ultra subscribers.
  • Gemini app: paid subscribers can generate longer, customized tracks for personal videos, podcasts or tutorials.
  • ProducerAI: this collaborative music creation tool is adding Lyria 3 Pro to help artists, producers and songwriters try out and revise complete songs. It is available worldwide to free and paid users.

What this means for you

If you use AI to create content, the most practical change is that you can request a piece with a more precise structure and adapt it more effectively to your project. For example, someone preparing a tutorial can generate a track with a short intro, a less intense central section and an outro. A video creator can request a longer song without having to manually join several fragments.

For developers, the arrival of Lyria on Vertex AI and the Gemini API opens the door to integrating music generation into their own applications. In that case, the value is not just creating one song, but producing many customized tracks for video games, videos or creative platforms.

Limits and safeguards

Google says it is developing Lyria alongside musicians, producers and songwriters. It also says the model does not directly imitate artists: when a user mentions a creator, the request is interpreted as general inspiration.

The company says it trains Lyria 3 and Lyria 3 Pro using materials that Google and YouTube have the right to use under their terms, agreements with partners and applicable law. It also applies filters to compare results with existing content and prohibits uses that violate copyright or privacy rights.

All Lyria 3 and Lyria 3 Pro results include SynthID, an imperceptible watermark that helps identify AI-generated content from Google.

The next step will be seeing how far these tools can go in creating usable music, rather than just quick ideas. Three-minute tracks expand the possibilities, but the real difference will come down to control over structure, consistency of the output and how well the music fits into the work of creators and professionals.

Google expands Lyria 3 Pro to create longer tracks | neversleep.ai