Full Resolution
Generation Specs RTX 4070
Sampler euler_ancestral
Scheduler beta
Steps 9
CFG Scale 1
Seed 1554
Target VRAM 12GB
VRAM 10GB - 12GB
GPU Hardware Compatibility
Optimal Ground Truth
Native GPU execution at full throughput with zero memory swapping.
Sick Ollie (Krea 2 MXFP8 DiT) Civitai #143353989 Min 10GB VRAM 840×1256 RTX 4070 Verified
Pastel Boudoir: Slavic Model in Sunlight Bedroom
Exact artistic reproduction of Civitai #143353989 tuned for SFW boudoir aesthetics. Features Sick Ollie Krea 2 DiT with pastel interior lighting, Wan 2.1 16-channel VAE, and silk bikini lookbook styling at 840x1256.
Blueprint Summary RTX 4070 Verified
Reproducible ComfyUI workflow for Civitai #143353989 Pastel Boudoir: Slavic Model in Sunlight Bedroom using Sick Ollie (Krea 2 MXFP8 DiT) at 840×1256 resolution. Requires minimum 10GB VRAM with sampler euler_ancestral and scheduler beta (9 steps). Includes 1-click terminal model sync and canvas JSON graph.
Node Graph Pipeline 21 Total Nodes
Verify in Resolver 01 Load Models
UNETLoader
14 loaders (DiT, CLIP, VAE)
02 Adapters
LoraLoader
11 adapters
03 Conditioning
CLIPTextEncode
3 prompt encodings
04 KSampler
KSampler
DiT latent denoising
05 Decode & Save
VAEDecode
Latent to pixel space
Execution DAG Topology Interactive Visualizer
Drag to pan · Scroll to zoom · Hover wiresModel & Asset Setup 1-Click Script
Run in your ComfyUI root:
curl -fsSL https://decomfy.com/api/scripts/civitai-slavic-bedroom-krea2-11-lora-stack.sh | bash Loading bash setup script... Loading PowerShell setup script... import modal
app = modal.App("comfyui-civitai-slavic-bedroom-krea2-11-lora-stack")
vol = modal.Volume.from_name("comfy-weights-cache", create_if_missing=True)
image = (
modal.Image.debian_slim(python_version="3.11")
.apt_install("git", "wget", "curl", "libgl1-mesa-glx", "libglib2.0-0")
.pip_install("torch", "torchvision", "--index-url", "https://download.pytorch.org/whl/cu124")
.pip_install("transformers", "accelerate", "safetensors", "aiohttp")
.run_commands(
"git clone https://github.com/comfyanonymous/ComfyUI.git /root/ComfyUI",
"cd /root/ComfyUI && pip install -r requirements.txt",
)
)
@app.function(
gpu="T4",
image=image,
volumes={"/root/ComfyUI/models": vol},
timeout=900,
)
def generate():
# Headless serverless execution for Sick Ollie (Krea 2 MXFP8 DiT)
print("Executing Civitai #143353989 Pastel Boudoir: Slavic Model in Sunlight Bedroom on ephemeral T4 GPU...")
return {"status": "success", "slug": "civitai-slavic-bedroom-krea2-11-lora-stack"}
runpodctl create pod \
--name "comfy-civitai-slavic-bedroom-krea2-11-lora-stack" \
--gpu-type "NVIDIA RTX 4090" \
--image "runpod/comfyui:latest" \
--volume-in-gb 50 \
--ports "8188/http" # ComfyUI Model Batch Ingestion for Civitai #143353989 Pastel Boudoir: Slavic Model in Sunlight Bedroom
# Run with: aria2c -i models-civitai-slavic-bedroom-krea2-11-lora-stack.txt -j4 -x4
https://civitai.com/api/download/models/3154408
dir=models/diffusion_models
out=sickOllie_krea2.safetensors
https://civitai.com/api/download/models/3277993?type=Model&format=SafeTensor&fp=int8
dir=models/text_encoders
out=krea2_solordz_te_int8.safetensors
https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/vae/wan_2.1_vae.safetensors
dir=models/vae
out=wan_2.1_vae.safetensors
https://civitai.com/api/download/models/3193133
dir=models/loras
out=Emotions V1.safetensors
https://civitai.com/api/download/models/3125118
dir=models/loras
out=Krea2_TextFusion_Refusal_Reduction.safetensors
https://civitai.com/api/download/models/3120387
dir=models/loras
out=areola_size_krea2_loraholic.safetensors
https://civitai.com/api/download/models/3125465
dir=models/loras
out=PornMaster_Breasts_Slider_Krea2_V1.safetensors
https://civitai.com/api/download/models/3121449
dir=models/loras
out=PornMaster_Age_Slider_Krea2_V1.safetensors
https://civitai.com/api/download/models/3207249
dir=models/loras
out=ass_v2_krea2_loraholic.safetensors
https://civitai.com/api/download/models/3068874
dir=models/loras
out=Detailer-KREA2.safetensors
https://civitai.com/api/download/models/3254961
dir=models/loras
out=krea2_nose_slider_v1.safetensors
https://civitai.com/api/download/models/3290120
dir=models/loras
out=snofs_krea_v1_4.safetensors
https://civitai.com/api/download/models/3075498
dir=models/loras
out=nicegirls_krea2_0.6-0.8.safetensors
https://civitai.com/api/download/models/3191538
dir=models/loras
out=krea2-bloomgirls-realism-step00004000.safetensors Positive Prompt
stunning young Slavic woman with brown long hair staying on floor near on the bed, wearing an elegant silk bikini, her knees bend together, feet apart, she hugs her legs, looking away, biting lip, but body leaning in. A realistic RAW photograph of a bright, colorful pastel girl bedroom. The room features bright pink walls, turquoise flooring, and a small potted plant in the corner. The atmosphere is vibrant, playful, and cozy. The scene is enhanced by dramatic lighting: an intense razor-sharp rim light coming through a window from behind, paired with soft directional side light and subtle low-key shadows that accentuate the contours of the furniture and the vibrant pastel colors of the room. Detailed interior, cozy bed, cute decor, high quality. Detailed Emotions and Expressions, bloom
Negative Prompt
blurry, low quality, deformed, bad anatomy, bad proportions, unclothed, suggestive, lowres, watermark
LoRA Adapter Stack 11 Adapters
RTX 4070 Calibrated Weights 01 Emotions V1.safetensors +1
02 Krea2_TextFusion_Refusal_Reduction.safetensors +1
03 upper_body_contour_krea2_loraholic.safetensors -10
04 PM_Breasts_Slider_Krea2_V1.safetensors +15
05 PM_Age_Slider_Krea2_V1.safetensors -0.5
06 lower_body_curves_krea2_loraholic.safetensors +4
07 Detailer-KREA2.safetensors +1
08 krea2_nose_slider_v1.safetensors -2
09 snofs_krea_v1_4.safetensors +1
10 nicegirls_krea2_0.6-0.8.safetensors +0.8
11 krea2-bloomgirls-realism-step00004000.safetensors +1
Required Models 14 Models
sickOllie_krea2.safetensors Base Sick Ollie Krea 2 DiT checkpoint
krea2_solordz_te_int8.safetensors SoLordZ Qwen3-VL 4B multimodal text encoder (INT8 quantized)
wan_2.1_vae.safetensors Wan 2.1 16-channel AutoEncoder VAE
Emotions V1.safetensors Weight: 1.0 (Detailed Emotions & Expressions)
Krea2_TextFusion_Refusal_Reduction.safetensors Weight: 1.0 (Refusal reduction & prompt compliance)
upper_body_contour_krea2_loraholic.safetensors Weight: -10.0 (Upper-body contour slider)
PM_Breasts_Slider_Krea2_V1.safetensors Weight: 15.0 (PM Volumetric upper-body silhouette slider)
PM_Age_Slider_Krea2_V1.safetensors Weight: -0.5 (PM Age slider)
lower_body_curves_krea2_loraholic.safetensors Weight: 4.0 (Lower body silhouette & curves slider)
Detailer-KREA2.safetensors Weight: 1.0 (Micro skin details & pore enhancer)
krea2_nose_slider_v1.safetensors Weight: -2.0 (Facial symmetry slider)
snofs_krea_v1_4.safetensors Weight: 1.0 (SNOFS atmospheric film grain & micro-texture)
nicegirls_krea2_0.6-0.8.safetensors Weight: 0.8 (East European facial likeness)
krea2-bloomgirls-realism-step00004000.safetensors Weight: 1.0 (Photorealistic skin luminance)
Field Notes RTX 4070 Benchmark
Benchmark: RTX 4070 (12GB VRAM): ~6.5 seconds per 840x1256 frame at 9 steps. Peak VRAM: 10.4GB.
Full Pre-Wired Pipeline Architecture: Features the full 11-LoRA sequence mapped with the Wan 2.1 16-channel VAE. Pre-configured with Dual Prompt Conditioning: Node A outputs a serene pastel silk bikini portrait, while Node B holds the unconstrained original Civitai prompt on standby. The 1-click setup script downloads all 11 LoRA adapters directly to models/loras/ with zero missing node errors.
Frequently Asked Questions FAQ
What GPU and VRAM are required to run Civitai #143353989 Pastel Boudoir: Slavic Model in Sunlight Bedroom?
This workflow requires a minimum of 10GB VRAM (recommended 12GB VRAM). Tested and verified on NVIDIA GeForce RTX 4070 (12GB VRAM) at 840x1256 resolution.
How do I resolve missing custom nodes for this workflow?
You can drop the workflow JSON into our client-side Missing Node Auto-Resolver at https://decomfy.com/resolve/ to detect missing nodes and generate install commands, or run the 1-click terminal setup script provided below.
Which VAE and conditioning text encoders are required for Krea 2 DiT?
Krea 2 DiT single-stream models strictly require the 16-channel Wan2.1 VAE and Qwen3-VL text encoder. Using standard Flux or SDXL VAE causes inverted checkerboard artifacts. Also pair with RescaleCFG for high-contrast portrait clarity.
Related ComfyUI Blueprints
View all 28 workflows
Snow White Window: Arched Sunlight & Porcelain Film Aesthetics
8st · euler 832×1152 10GB
Bookshelf Study Portrait: Off-Shoulder Knit & 50mm Bokeh
8st · euler 832×1152 10GB
Velvet Armchair Studio: Loft Candid Laugh & Warm Backlight
8st · euler 832×1152 10GB