Lyria

Lyria is Google DeepMind's high-fidelity, text-to-music generative AI model, designed for creating professional-grade audio and compositions from simple text prompts.

Lyria is a powerful generative AI model from Google DeepMind, focused on high-quality music creation. The latest iteration, Lyria 2, produces professional-grade 48kHz stereo audio, capturing intricate compositions across diverse genres. It is accessible to developers via the Gemini API and powers tools like YouTube's Dream Track. A key feature is Lyria RealTime, which enables interactive, continuous music streaming and control. For responsible deployment, Lyria integrates SynthID, an imperceptible digital watermarking technology that identifies the audio as AI-generated content.

https://deepmind.google/models/lyria/