6 tips for prompting Lyria 3 in the Gemini app
This guide offers six tips for effectively prompting Google DeepMind's Lyria 3 generative music model within the Gemini app.Summary
Lyria 3, Google DeepMind’s latest generative music model available in the Gemini app, allows users to create original 30-second tracks using text, images, or videos. To maximize results, users should begin by writing detailed text prompts, specifying elements like theme, genre, and desired vocal style. A second tip is to upload visuals like photos or videos as inspiration, allowing Lyria to generate a matching musical score based on the image content. Users are encouraged to define the genre and era, even blending styles like K-pop with Motown, or to add specific details regarding instruments, song dynamics (e.g., quiet build-up to an explosive chorus), and vocal characteristics (e.g., rich, gravelly, or layered harmonies). For lyrics, users can either input their own using the "Lyrics:" code or ask Lyria to generate them based on a clear theme. Finally, created tracks can be easily shared by downloading them as MP3 or MP4 files.
(Source：Gemini)