Google unveils Lyria 3 Pro for longer, structured AI music generation

New AI music model enables longer, structured compositions with improved creative control, expanding across Google platforms including Vertex AI, Gemini, and Vids to support developers, businesses, and content creators globally.

Google has introduced Lyria 3 Pro, the latest version of its AI-powered music generation model, as part of efforts to enhance creative capabilities across its ecosystem. Developed by Google DeepMind, the upgraded model is designed to produce longer and more structured music tracks, catering to a wide range of professional and creative use cases.

Enhanced composition and creative control

Lyria 3 Pro brings significant improvements in music composition, enabling users to generate tracks of up to three minutes. Unlike earlier versions, the model allows more precise control over musical structure, supporting prompts that define elements such as intros, verses, choruses, and bridges. This enables creators to build complete compositions with smoother transitions and consistent flow.

The system is designed to better understand musical patterns, offering flexibility to experiment across genres and styles. It is particularly suited for applications such as video production, podcasts, and digital content, where customised background music is increasingly in demand.

Wider integration across platforms

Google is rolling out Lyria 3 Pro across multiple platforms, including Vertex AI for enterprise-scale use, as well as Google AI Studio and the Gemini API for developers. The model is also being integrated into Google Vids, allowing users to generate music directly within video projects, while Gemini app users will gain access to enhanced music creation features.

The company emphasised responsible AI use, incorporating safeguards such as SynthID watermarking and restrictions on mimicking specific artists, while continuing to expand its AI-driven creative ecosystem.