Google's Lyria 3 Pro powers longer AI music

Google DeepMind is enhancing its AI music generation capabilities with the introduction of Lyria 3 Pro. This advanced model builds upon the earlier Lyria 3, enabling the creation of significantly longer musical pieces, now extending up to three minutes. This expansion in track length is coupled with a deeper understanding of musical structure.

Lyria 3 Pro allows users to prompt for specific song components, such as intros, verses, choruses, and bridges, offering greater creative control and facilitating the generation of songs with complex transitions. This makes it a powerful tool for experimentation and detailed composition.

The capabilities of Lyria 3 Pro are being integrated across Google's product ecosystem. Businesses can access the model for on-demand audio at scale through Vertex AI's public preview. Developers will find Lyria 3 Pro available in Google AI Studio and via the Gemini API, enhancing creative tools with improved musicality and structural coherence.

For everyday users, Lyria 3 Pro is coming to Google Vids, an AI-powered video creation app, allowing for custom soundtracks to match video content. Paid subscribers of the Gemini app will also gain access to longer generation capabilities. Additionally, the model is powering ProducerAI, a collaborative music creation tool for artists and producers.

Google continues to develop its music AI responsibly, partnering with industry professionals. Insights from their Google DeepMind music generation efforts, including the Music AI Sandbox, have informed Lyria 3's development. The company emphasizes that Lyria 3 and Gemini do not mimic specific artists and employ filters to prevent infringement, with all outputs watermarked using SynthID.