Live2D model used in the Iroha speech demo.
Model source: A8Chann/dsh-pet-live2d.
Artwork and model: CC BY-NC-SA 4.0. Original license and attribution.
The original model files are unchanged. This site's integration adds audio-driven lip sync and maps the existing speech emotions to the model's expressions. This fan-created character is not an official DeepSeek mascot.
Renderer: PixiJS 8.19.0 and untitled-pixi-live2d-engine 1.3.5, bundled by dsh-pet-live2d under the MIT license. Bundle license; PixiJS license; engine license; tiny-lru license.