Kandinsky LabKandinsky 2
Kandinsky 2.2 Image Prior
This trained prior maps a text condition into CLIP image embeddings. It belongs to the 2.2 pipeline and works with its separate image decoder.
Match the prior and image encoder generation
The release uses CLIP-ViT-G representations and a retrained prior. Swapping in the 2.1 prior is not the documented 2.2 pipeline.kandinsky-community/kandinsky-2-2-prior — model card and files
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | kandinsky-community/kandinsky-2-2-prior |
| Weight license | Apache 2.0kandinsky-community/kandinsky-2-2-prior — model card and files |
Related models
- Kandinsky 2.2
Kandinsky Lab · Kandinsky 2
A 1024-pixel image model with a retrained CLIP-based prior and an adapted diffusion decoder.
Launch date not confirmedDownloadable