Google officially launched Lyria 3 Pro this Wednesday, a significant upgrade to its music generation model that extends track lengths from 30 seconds to three full minutes. This release comes just one month after the debut of the standard Lyria 3 model, marking a rapid acceleration in Google’s generative audio capabilities.
Enhanced Creative Control and Structural Intelligence
Beyond the extended duration, Lyria 3 Pro introduces advanced customization features. The model demonstrates a deeper understanding of musical architecture, allowing users to specify distinct components—such as intros, verses, choruses, and bridges—within their prompts. This structural awareness provides creators with significantly more control over the final output compared to its predecessor.
Integration Across Gemini and Enterprise Platforms
The Pro model is currently rolling out across several key Google products:
- Gemini App: Now available for paid subscribers.
- Google Vids: Integrated into the company’s video editing software.
- ProducerAI: The model is being deployed within this GenAI-powered production tool, which Google acquired last month.
For enterprise users, Google has introduced Lyria 3 Pro to Vertex AI (currently in public preview), the Gemini API, and AI Studio, expanding the reach of its music generation technology for developers and businesses.
Training Standards and Ethical Safeguards
Google clarified that the model was trained using data from authorized partners, alongside permissible data from YouTube and its own internal archives. Addressing concerns regarding artist identity, the company stated that the model does not mimic specific performers. When users reference an artist in a prompt, the system claims to draw only “broad inspiration” from that style.
To ensure transparency, all audio generated by both Lyria 3 and Lyria 3 Pro is embedded with SynthID, a digital watermark that identifies the content as AI-generated.
The Evolving AI Music Landscape
The release arrives amid heightened industry scrutiny regarding AI-generated audio. Earlier this week, Spotify introduced new tools enabling artists to verify tracks released under their names to prevent misattribution by “AI slop” creators. Simultaneously, Deezer has deployed detection tools designed to help streaming platforms identify and label AI-produced music, highlighting the growing need for accountability in the generative media sector.