Google’s Gemini App Now Creates Music with Lyria 3

Google has integrated its most advanced AI music generation model, Lyria 3, directly into the Gemini application. This modern feature empowers users to generate 30-second music tracks from simple text prompts or image uploads [1].

How Lyria 3 Works

Lyria 3 allows users to create custom music by describing their desired sound – including genre, mood, or rhythm – or by uploading a photo. The AI then generates a complete song, potentially including vocals and lyrics [1]. The system can also analyze the visual atmosphere of uploaded images or videos to create fitting background music.

Key Features and Improvements

• Advanced Music Generation: Lyria 3 represents a significant improvement over previous models, offering enhanced capabilities in generating lyrics, providing greater creative control, and producing more complex musical arrangements [1].
• SynthID Watermarking: All music generated within the Gemini app is embedded with SynthID, Google’s imperceptible watermark, to identify AI-generated content [1].
• Custom Cover Art: Generated tracks include custom cover art [1].
• Accessibility: The feature is available within the Gemini app for users aged 18 and above, in multiple languages, with increased usage limits for subscribers [1].

DeepMind’s Role and Musicality

Developed by Google DeepMind, Lyria 3 was created with input from music producers and musicians to ensure a strong understanding of musicality, including rhythm and harmony [2]. The model is designed to be a collaborative tool for musical exploration [3].

Expanding AI Music Creation

Google is also extending this technology to YouTube creators through the Dream Track feature, enabling automated soundtrack creation for Shorts videos [1].

Availability

The music generation feature powered by Lyria 3 is currently in beta and available within the Gemini app [4].