⚠️ IMPORTANT — Applies to v1.0 and v1.1 ⚠️
❌ Do NOT use the INT8 Fast custom node.
This checkpoint is already in native ComfyUI INT8 (ConvRot) format.
✅ Use the standard ComfyUI Load Diffusion Model node.
Requirements
- ComfyUI v0.27+
- ComfyUI Kitchen
Krea2 Center v1: similar, experimenting with some block merging.
Krea2 Center v0.5 explores whether Krea2 can become more expressive through carefully balanced LoRA merging while preserving the model's realism, flexibility, and overall character. The goal isn't to replace Krea, but to gently extend its emotional range across portraits, fantasy, sci-fi, and stylized scenes.
Added Int8: about 25% faster on 50 series Nvidia don't have a 40 or 30 series to test.
Krea 2 is licensed under the Krea 2 Community License Agreement. For more information, visit https://krea.ai/krea-2-licensing.
Description
Int8 works with native comfy loader
FAQ
Comments (8)
NVFP4... so we cant run it on AMD gpus?
I will see an upload an fp8 version
Bro thank you so much for this model, I love it! INT8 version is very fast and high quality. I am enjoying the more fantasy and sc-fi results
Big thanks for this model. Currently the best Turbo version among the few tunes/merges we have. Will you consider releasing the base/raw version, too? The Turbo is good, but the base/raw version has significantly more creativity and diversity for the same prompt. With the Turbo, you're usually stuck with variations of the same image regardless of seed, but the raw version delivers different generations across seeds.
maybe there will be a "semi-raw" version tomorrow getting good variance on the new merge
This is pretty amazing all by itself without any additional loras!
Have to agree, can get it to do most of the things i want, just takes some getting used to having to switch back to narrative prompting instead of tags. I think i will be switching from illustrious to Krea now because.. damn!
Well done, you can do crazy stuff with this checkpoint. Its awesome :) I will post further pictures to the gallery