Google Lyria 3 Pro AI music model arrive with longer music generation capability, more control

Google has introduced Lyria 3 Pro, an upgraded AI music model that can generate tracks up to three minutes long. The tool is available in public preview on Vertex AI, Google AI Studio, Google Vids, and the Gemini app for paid users.

Google has introduced a new AI tool called Lyria 3 Pro, which arrives as an upgrade to the existing Lyria 3, launched last month. The new model allows users to create music tracks that are up to three minutes long, a significant improvement over Lyria 3, which could generate tracks of up to 30 seconds.

Lyria 3 Pro also brings more customisation and creative control. Google says the model has a better understanding of musical composition, allowing users to prompt for specific song elements such as intros, verses, choruses, and bridges.

“It’s great for experimenting with different styles or generating songs with complex transitions,” Google wrote in a blog post.

Lyria 3 Pro is already available in public preview on several Google platforms. Businesses can access it through Vertex AI, while developers can use it via Google AI Studio. The tool is also available on Google Vids. In addition, paid users of the Gemini app can also access the tool.

Both Lyria 3 and Lyria 3 Pro can generate music tracks from text prompts, photos, and videos. To create a track using text, users can describe a specific genre, mood, inside joke, or memory. The model then generates a track with lyrics or instrumental audio that fits the prompt. Users can also generate tracks from photos or videos. By uploading a photo or video, Gemini can analyse the content and compose a track with lyrics that match the mood.

Google says the model was trained using materials from YouTube and other content that Google has the rights to use. The company also noted that music generated by Lyria 3 Pro is embedded with SynthID, which helps identify AI-generated content.