Lyria (Google)
Web3 / ai data
Lyria is Google DeepMind's core multimodal foundation model for professional high-fidelity music generation. It translates text, image, or video inputs into professional-grade musical tracks with precise control over tempo, genre, and emotional mood. The current release, Lyria 3, generates tracks up to three minutes long and can produce highly realistic AI vocal performances in multiple languages, accompanied by structurally coherent automated lyric writing and complex instrumental arrangements. All Lyria-generated content is embedded with an imperceptible SynthID watermark for AI provenance identification. Lyria powers the Lyria RealTime engine used by MusicFX DJ for live composition, and is integrated into Google Flow for synchronised audio-video generation alongside Veo. Example: A token project generates its full brand audio identity using Lyria 3 — a 3-minute ambient theme for the dApp, a shorter energetic sting for social ads, and a vocal jingle in English, Spanish, and Korean — all consistent in style, all SynthID-watermarked, and all produced without a music producer or licensing cost. Why it matters for AI and data in Web3: Original music for Web3 products — games, virtual worlds, promotional content — traditionally required either expensive commissioned composers or navigating music licensing rights. Lyria 3 makes bespoke, professional-quality music generation free and instantaneous, with built-in AI watermarking relevant to on-chain provenance frameworks.
Explore the full Web3 Glossary — 2,062+ expert-curated definitions. Need guidance? Talk to our consultants.