Google has launched Lyria 3, its latest generative music model, in the Gemini app. Users can create 30-second custom tracks with lyrics and cover art by describing a prompt in text or uploading a photo/video. Key improvements over previous Lyria models include automatic lyric generation, greater creative control over style and tempo, and more realistic, musically complex output. All generated tracks are watermarked with SynthID for AI content verification. The feature is available in beta for users 18+ in multiple languages, with higher usage limits for paid Google AI subscribers. Lyria 3 is also expanding to YouTube Dream Track for Shorts creators.
Table of contents
New audio verification capabilitiesOur commitment to developing generative AI responsiblySort: