- Google's Lyria 3 Pro launches for Gemini app paid subscribers, enabling AI-generated music tracks up to three minutes in length.
- The enhanced model demonstrates sophisticated understanding of musical composition principles and processes intricate, multi-layered prompts.
- Lyria 3 suite deployment begins this week across multiple platforms: Gemini app, Vertex AI, Google AI Studio, Gemini API, Google Vids, and ProducerAI.
Google's Gemini app leverages a diverse ecosystem of specialized AI models for multimodal content creation—Nano Banana handles image synthesis, Veo powers video generation, and Lyria drives music production. Last month saw the integration of Lyria 3 into the Gemini platform, though initial capabilities were limited to brief 30-second audio clips. Today marks a significant expansion of these capabilities with the introduction of Lyria 3 Pro. This advanced iteration delivers substantially improved comprehension of musical structure and composition while extending track duration sixfold. The rollout to Gemini app subscribers begins immediately, Google confirmed in today's blog post.
The enhanced musical intelligence of Lyria 3 Pro enables the model to parse granular, technically specific prompts that incorporate professional music terminology. This advancement allows users to provide precise instructions for individual song components rather than broad, holistic descriptions. When generating output, Lyria 3 Pro evaluates these element-specific directives independently, giving creators fine-grained control over how particular sections, instruments, or musical phrases should be rendered within the composition.
