Google launches Lyria 3 Pro music generation model

In an exciting development for music enthusiasts and creators, Google has introduced the Lyria 3 Pro music generation model, just a month after the launch of its predecessor, Lyria 3. This new model empowers users to craft tracks that can extend up to three minutes in length, a significant upgrade from the previous limit of 30 seconds. Lyria 3 Pro not only allows for longer compositions but also enhances creative control and customization. Users can now specify various components of their musical creations, such as intros, verses, choruses, and bridges, thanks to the model's improved understanding of track structures. This innovative model will be integrated into the Gemini app, although access will be limited to paid subscribers. Additionally, Google plans to incorporate Lyria 3 Pro into its Google Vids video editing application and ProducerAI, a recently acquired AI-powered music production tool. Moreover, Google is expanding music generation capabilities across its enterprise tools, including Vertex AI in public preview, the Gemini API, and AI Studio, utilizing the advancements from Lyria 3 Pro. The company has clarified that the model was trained using data from its partners, as well as permissible data from platforms like YouTube and Google, ensuring compliance and originality. Importantly, while Lyria 3 Pro can draw 'broad inspiration' from specified artists, it does not replicate their work. Each track generated using Lyria 3 and Lyria 3 Pro will carry a SynthID label, indicating it was created with AI assistance. In a related move, Spotify recently introduced tools for artists to oversee songs released in their name, aiming to prevent misattribution by AI-generated content. Similarly, Deezer has rolled out features to help streaming services identify AI-generated music.