Google has launched Lyria 3, its AI music generation model, within the Gemini app. This new feature allows users to create 30-second songs from text prompts or uploaded images, complete with lyrics and instrumental backing.
While the model produces coherent tracks within mainstream genres like pop and R&B, it struggles with more niche styles and a strict 30-second output cap. This limitation positions it behind competitors such as Suno and Udio, which offer longer song generation, deeper creative controls, and more mature workflows.
Google emphasizes its commitment to copyright, stating Lyria 3 avoids mimicking specific artists, using names only for mood inspiration. Tracks are embedded with SynthID watermarks for provenance, and Gemini offers audio verification to identify AI-generated content.
Lyria 3 is currently available on desktop for Gemini users 18 and older, with mobile rollout imminent. Premium subscribers receive higher generation limits. The technology is also expanding YouTube's Dream Track feature globally, providing Shorts creators with AI-generated soundtracks.