Interactive Layout Node & Prompt Generator for MOSS-Image
Eliminate manual [x1, y1, x2, y2] coordinate guessing. Draw spatial layout regions visually on the calibrated grid to generate ComfyUI node parameters and exportable workflow JSON instantly.
Dictates holistic tone, ambient illumination, depth, and background cohesion.
Generated ComfyUI Node Payload & Coordinates
Paste directly into MOSS_Image_Layout_Conditioning or export JSON
# ComfyUI MOSS-Image Layout Conditioning Node Values
# Resolution: 1024x1024 | Box Count: 3
layout_boxes = [
[0.117, 0.078, 0.5, 0.41], # Neon Billboard Region
[0.469, 0.352, 0.938, 0.957], # Mecha Girl Focal Character
[0.078, 0.732, 0.449, 0.938], # Ground Reflections & Hover Car
]
prompts = [
"Glow text \"MOSS\" written in bold Kanji and English neon tubes, vibrant turquoise and magenta halo",
"Full body anime cyborg warrior, white ceramic armor with carbon accents, holding glowing Katana, wet hair",
"Vintage futuristic hover car parked on wet ground, chrome headlights casting long puddle refractions"
]
weights = [1.00, 1.20, 0.95]
global_prompt = "Cinematic wide angle, rain-slicked Neo-Tokyo alleyway, 8k resolution, volumetric neon lighting, photorealistic reflections"
Rendered Latents from Exact Same Multi-Region Layout
Notice how each bounding box enforces strict spatial containment without prompt bleeding.
Luminescent signage contained within strict boundary without bleeding into street geometry.
Cyborg anatomy is pinned to the right quadrant, holding glowing Katana without interference.
Complete tensor output merging ambient alleyway lighting with all 3 spatial conditioning boxes.
How to Wire the Layout Node in ComfyUI: 3-Step Setup
Connect directly to MOSS_Image_Layout_Conditioning without heavy auxiliary ControlNets.
Add Layout Node
Right-click on ComfyUI canvas → Add Node → conditioning → MOSS_Image_Layout_Conditioning.
Paste Normalized Coordinates
Click [Copy Raw Coordinates] above and paste into the node’s layout_boxes parameter widget.
Connect to KSampler
Wire the CONDITIONING output from the Layout node straight into the positive conditioning slot of your KSampler.
⚡ Run MOSS-Image Online (Zero Install, Zero VRAM Headache)
Bypass the 15GB model downloads and CUDA memory errors. Launch an RTX 4090 cloud ComfyUI instance pre-loaded with MOSS-Image INT8/BF16 weights.