AI Potluck
Back to Gap Map Model components / Image, video, 3D & music generation

Lyria

Google
closed / Overall score: n/a

Lyria is Google DeepMind's music generation model line, which produces full-length songs with vocals and instrumental clips from text prompts. It is served through the Gemini API and in Google products including Gemini, YouTube Shorts and Flow, alongside a real-time variant for interactive generation.

Magenta RealTime, which Google describes as its open Lyria model, is a separate open-weight release and is not part of this entry.

Openness

1 high confidence
1.0
weights
closed(no weights are distributed
data
closed
code
closed
license
Proprietary(Gemini API terms, which bar attempts to extract or replicate the model)

Lyria is available only through Google's API and apps, and Google's terms forbid extracting the model. Google has released a separate, smaller open model, Magenta RealTime, but no Lyria weights, data or code.

Adoption

not assessed

Google publishes no usage figure for Lyria, and a hosted model has no download count, so no reading was possible.

Capability

4 medium confidence
4.0

Lyria places just below the top three on the instrumental music arena and above every downloadable model. Suno holds the two places above it on both boards.

Verified 2026-09-26