
Lyria 3 by Google Deepmind
Turn any photo or thought into a custom song inside Gemini
Lyria 3 by Google Deepmind adds a music generation model to Gemini, allowing users to produce short tracks from text prompts or images. By describing a mood, uploading a picture, or entering an idea, the system creates a 30‑second song with vocals and generates matching cover art. The feature works both in the app and on the web, providing an immediate way to turn visual or textual input into custom audio.