gguf-org/krea-2-gguf

krea-2-gguf

• drag pig to > ./ComfyUI/models/vae
• drag qwen3vl-4b to > ./ComfyUI/models/text_encoders
  • compatible with this, this, this, etc. (but 🐷 reversible, opt anyone you like)
• drag krea2 to > ./ComfyUI/models/diffusion_models
• 

Prompt

cute anime style girl with massive fluffy fennec ears and a big fluffy tail blonde messy long hair blue eyes wearing a maid outfit with a long black gold leaf pattern dress and a white apron, it is a postcard held by a hand in front of a beautiful realistic city at sunset and there is cursive writing that says "Now in GGUF"

Prompt

A crystal dragon soaring through an aurora borealis sky, its entire body made of transparent faceted crystal refracting the green and purple aurora light into rainbow spectra, ice particles trailing from its wings, high fantasy digital art

Prompt

A crystal phoenix soaring through an aurora borealis sky, its entire body made of transparent faceted crystal refracting the green and purple aurora light into rainbow spectra, ice particles trailing from its wings, high fantasy digital art

Prompt

iq1_m (same prompt as 1st one)

Prompt

iq2_xxs (same prompt as 1st one)

Prompt

iq2_xs (same prompt as 1st one)

Prompt

mxfp4 (same prompt as 1st one)

Prompt

nvfp4 (same prompt as 1st one)

Prompt

nvfp4 (model) + nvfp4 (clip)

compatible with gguf-desktop or stable-diffusion.cpp

note: feel free to test nvfp4 and mxfp4 on gguf node and/or gguf-desktop ⚽🐖 *for initial test, iq3-bit quants might not work good in the node but it works pretty good in gguf-desktop, however, iq2-bit quants work better in gguf node than gguf-desktop, the discrepencies might due to the way how the engine handle those tensors and varies by models, btw, you could make all the other quant types yourself with Quant tab from gguf-desktop and test it one by one; enjoy your long weekend