Google's New AI Makes 3-Minute Songs, Cuts Memory 6x

Google just made artificial intelligence more creative and more efficient with two major announcements that could change how we interact with technology.

The tech giant unveiled Lyria 3 Pro, a music generation tool that lets anyone create full songs up to three minutes long. Users can now structure their compositions with intros, verses, and choruses, giving them real creative control over their musical ideas.

The tool is rolling out across Google's platforms, including the Gemini app, Vertex AI, and AI Studio. It's also being integrated into Google Vids and ProducerAI, though some features will be limited to paid subscribers.

What makes this exciting is how it democratizes music creation. You don't need years of training or expensive equipment anymore. If you have a melody in your head or a song idea you want to explore, you can bring it to life in minutes.

But Google didn't stop there. The company's research team also introduced TurboQuant, a memory compression algorithm that makes AI systems run six times more efficiently without losing quality.

This technical breakthrough tackles one of AI's biggest challenges: the massive computing power needed to run these systems. By shrinking memory requirements through advanced compression, TurboQuant could dramatically lower the cost of running AI applications.

The Ripple Effect

These advances matter beyond just Google's ecosystem. Cheaper AI infrastructure means smaller companies and researchers can access powerful tools without breaking the bank. When technology becomes more affordable and accessible, innovation spreads faster.

The music generation tool opens creative possibilities for content creators, educators, and hobbyists who previously couldn't afford professional production. Meanwhile, TurboQuant's efficiency gains could accelerate AI development across industries, from healthcare to education.

Google Research will present the TurboQuant findings at ICLR 2026, one of the world's leading AI conferences. The fact that they're sharing this research openly means other organizations can build on these breakthroughs.

Together, these announcements show AI evolving in two crucial directions: becoming more helpful for everyday creativity and more sustainable to operate at scale.