Moody RED - v1.0 Krea2

note : some of the images used krea realism lora

in this merge i used redcraft as base since it has really good contrast, colors, and anatomy, combined with Moody Krea2 Mix

the goal is to adjust redcraft to keep those same dynamic poses and energy, while the moody influence helps balance the fantasy style into something a bit more grounded and real. it's a block merge i used the default krea2 turbo workflow for generating the images in showcase
its should be able to do both nsfw and sfw

Try Moody Red On : Moescape - TensorArt - PIXAI

Model Differences (Open in new tab if text is too small)

Block Merge Breakdown:

• Redcraft (Base Model)

• Moody Krea2 Mix

Images in the Checkpoint cover image used setting below :

• Krea2 Turbo Workflow

• Sampler - Euler

• Steps - 8 Steps

• CFG Scale - CFG 1

Recommended Settings :

• Sampler - Euler / Euler-sde

• Steps : 6-12

• CFG : 1.0

How It Works & Prompting Tips :

since this model is built on the krea2 architecture, it handles prompts differently than old-school diffusion models. krea2 was trained natively to understand natural language, meaning you don't need to throw a messy wall of random tags and weight syntax at it.

instead, try writing your prompts using clear, descriptive sentences focusing on lighting, mood, materials, and atmosphere

Pro-Tip for Prompting : if you want the cleanest possible results out of this architecture, feel free to use LLMs like gpt, gemini, or deepseek to draft out rich, descriptive natural language prompts, or use the civitai prompt enhancer. it helps pull out way better details and lighting descriptions

Experiment different settings and tags to your Generating style and likings

Credits

Huge thanks to the amazing model creators:

• @AiMetatron

• @catlover1937

Dont forget to check out their original work as well.

Made with love ❤️