Google Lyria 2

AI model for music generation that produces high-fidelity audio output.

Model Overview

A sophisticated AI model designed for music generation, capable of producing high-fidelity 48kHz stereo audio based on text-based prompts.

Best At

Lyria 2 excels at generating a wide range of musical genres, including Classical, Jazz, Pop, Electronic, Orchestral, and various regional and fusion styles. It's ideal for creating professional-grade audio suitable for direct integration into projects.

Limitations / Not Good At

Currently, the model has restricted language support and may have limited genre coverage in certain niche musical styles. Performance can vary with different musical complexities.

Ideal Use Cases

• Creating background music for videos or podcasts
• Generating musical ideas for composers and producers
• Developing soundscapes for games or interactive media
• Experimenting with new musical styles and fusions

Input & Output Format

Input: Text prompt (string), optional negative prompt (string), optional seed (integer).
Output: 48kHz stereo audio (URI).

Performance Notes

Generates up to 30 seconds of audio. Implements SynthID for audio watermarking.