Lyria 2 is a music generation model that produces 48kHz stereo audio through text-based prompts
Frequently asked questions
What is Google Lyria 2 and what does it do?
Google Lyria 2 is an AI music generation model that converts natural language text prompts into 30 seconds of high-fidelity, 48kHz stereo audio. It handles melody, harmony, rhythm, and instrumentation automatically, making it suitable for creators who need professional-grade music without a recording studio or musical expertise.
What audio quality does Lyria 2 produce?
Lyria 2 produces stereo audio at 48kHz, the same sample rate used in professional music production and broadcast media. This means outputs are broadcast-ready and can be integrated directly into video timelines, game engines, or advertising projects without additional post-processing or audio engineering.
What music genres does Lyria 2 support?
Lyria 2 supports a broad range of musical styles including Classical, Jazz, Pop, Electronic, Orchestral, and various regional and fusion genres. Genre selection is controlled entirely through descriptive text prompts, giving creators flexible creative control without needing MIDI input, stems, or any audio file uploads.
What is SynthID watermarking in Lyria 2?
SynthID is Google's AI watermarking technology embedded in every audio file generated by Lyria 2. It provides transparent AI provenance, meaning the audio can be identified as AI-generated even after editing or compression. This supports responsible content use and helps creators and platforms comply with AI transparency standards.
What are the best use cases for Lyria 2?
Lyria 2 is best suited for background music in video content, rapid music prototyping, game development audio, advertising jingles, and any creative pipeline requiring polished audio quickly. Because it outputs professional stereo audio directly from text, it eliminates the need for composers or studio sessions for many common production scenarios.
Does Lyria 2 require audio input or MIDI files?
No. Lyria 2 operates exclusively in the audio domain using text prompts as the sole input. It does not require MIDI files, audio stems, or any existing audio input. All creative direction — including genre, mood, tempo, and instrumentation — is communicated through natural language descriptions in the prompt.
How do I use Lyria 2 on CRAISEE?
On CRAISEE, Lyria 2 is available via the Replicate provider under the Text-to-Audio category. Users write a descriptive text prompt specifying the desired musical style, mood, and instrumentation. The model then generates a 30-second, 48kHz stereo audio file ready for download and direct use in creative projects.
Who created Lyria 2 and where is it hosted?
Lyria 2 was created by Google and is hosted on the Replicate platform, accessible through CRAISEE. It is Google's flagship AI music generation model, designed for creators, composers, and developers who need high-quality, AI-generated music output integrated into professional creative and development workflows.
