Google’s Lyria 3 Pro Ushers in a New Era of AI-Powered Music Creation
Google has launched Lyria 3 Pro, an advanced music generation model building upon its predecessor, Lyria 3. This new iteration empowers users to create longer, more customizable musical tracks, signaling a significant step forward in the accessibility and sophistication of AI-driven music production. The launch expands Google’s AI music tools across its Gemini platform, enterprise products, and other services.
Longer Tracks and Enhanced Control
Lyria 3 Pro distinguishes itself from Lyria 3 by increasing the maximum track length to three minutes, a substantial leap from the previous 30-second limit. This extended duration, coupled with improved understanding of musical composition, allows users to specify elements like intros, verses, choruses, and bridges within their prompts, resulting in more structured and complete musical pieces. Google states that the model now better understands musical composition.
Accessibility and Integration
The Pro model is now available within the Gemini app, but access is currently restricted to paid subscribers. Beyond Gemini, Lyria 3 Pro is also being integrated into Google Vids, a video editing application, and ProducerAI, a GenAI-powered music production tool acquired by Google last month. TechCrunch reported on these integrations.
Enterprise Applications
Google is extending the capabilities of Lyria 3 Pro to its enterprise tools through Vertex AI (currently in public preview), the Gemini API, and AI Studio. This expansion aims to provide businesses with tools for creating custom audio content for a variety of applications.
Data and Artist Mimicry
Google emphasizes that Lyria 3 and Lyria 3 Pro were trained using data from its partners and permissibly sourced data from platforms like YouTube and Google. While the model is designed not to directly mimic artists, it can draw “broad inspiration” from specified artists when included in user prompts.
AI-Generated Music Identification
To ensure transparency, all tracks created using Lyria 3 and Lyria 3 Pro are marked with SynthID, a digital watermark indicating that AI was used in their creation. This addresses growing concerns about the identification of AI-generated content, particularly in the music industry, where Spotify and Deezer have recently launched tools to detect and attribute AI-generated songs.
Implications for Marketers and Creative Workflows
The advent of tools like Lyria 3 Pro signifies a shift in the landscape of audio content creation. Marketers can now experiment with creative ideation and campaign versioning by generating audio variations tailored to specific audiences, platforms, and moments. However, maintaining brand differentiation and originality will require human oversight to guide the AI and ensure the final audio remains distinctive and on-brand.
Availability
Lyria 3 Pro music generation in Gemini is currently available globally in English, Japanese, Korean, Hindi, Spanish, Portuguese, German, and French for users over the age of 18.
Worth a look