Skip to content
#

image-upscale-ai

Here are 4 public repositories matching this topic...

Seamless tiled refinement and upscaling for ComfyUI. The VL nodes step all tiles on one shared latent, conditioned by vision tokens and per-tile captions instead of a prompt, so objects never repeat across tiles. Refine whole images or just a masked region. Any GUIDER model on the base node; Krea 2 on the VL nodes. Tested at 8K+.

  • Updated Oct 1, 2026
  • Python

Add this topic to your repo

To associate your repository with the image-upscale-ai topic, visit your repo's landing page and select "manage topics."

Learn more