ComfyUI Layer Style
A set of nodes for ComfyUI it generate image like Adobe Photoshop's Layer Style. the Drop Shadow is first completed node, and follow-up work is in progress.
Nodes (169)
AutoAdjust
AutoAdjust V2 — auto white-balance and levels, with manual trim
AutoBrightness
Brightness & Contrast — a quick brightness, contrast and saturation tweak
Brightness Contrast V2 — the quick brightness / contrast / saturation knob
ColorAdapter — match one image's color to another's
ColorBalance
Color of Shadow & Highlight
Color of Shadow & Highlight V2 — split-tone your darks and lights separately
ColorTemperature
Exposure
Gamma
HSV Color Adjust (Layer Style)
LAB Color Adjust (Layer Style)
Levels — Photoshop Levels, per channel, in ComfyUI
LUT Apply
Negative
RGB
YUV
Add Grain
ChannelShake
ColorMap
Distort Displace
Film
Film V2 — grain, vignette and fake depth-of-field in one node
A dead-simple whole-image blur node
Gaussian Blur V2 — a blur node with a real float radius
Halftone (Layer Style)
HDR Effects — local-contrast punch without the crunchy halos (if you're careful)
Fake film light leaks over any image
Motion Blur (Layer Style)
Sharp & Soft
One-node skin smoothing that hands you the mask too
A dreamy glow/diffusion filter for highlights
Watercolor (Layer Style)
Blend If Mask (Layer Style)
CreateGradientMask
Generate a rounded-rectangle mask
Pull a mask out of any color channel, with levels
Load Segformer Model
Turn a mask into a bounding box and crop coords
Mask Box Extend
Mask by Color
Pull mask edges in to kill halos
Matting-grade edge refinement for rough masks
MaskEdgeUltraDetail V2 — rescue the hair your cutout mask threw away
MaskEdgeUltraDetail V3
Fade a mask off from one edge
Add noise to a mask edge for natural blends
MaskGrow
Flip a mask, black to white and back
Directional blur on a mask
MaskPreview — actually see the mask you're building
Make an outline mask from any mask edge
Bleed edge pixels outward to kill dark halos
RemBg Ultra (Layer Style)
RmBgUltra V2 — background removal with matting-grade edge cleanup
SegformerB2ClothesUltra — mask clothes and body parts with clean edges
SegformerClothesPipelineLoader — load the clothes-segmentation model once, reuse it
Segformer Clothes Setting
Segformer Fashion Pipeline
Pick which clothing parts to mask
Segformer Ultra V2
Segformer Ultra V3 — auto-mask clothes, faces and body parts by name
Shadow & Highlight Mask
Luminosity masks for shadows and highlights
Photoshop's Color Overlay layer style as a node
ColorOverlay V2
DropShadow — drop a proper shadow behind a cutout
DropShadow V2 — a real soft drop shadow behind your cutout
DropShadow V3 — the current drop-shadow node, background now optional
Gradient Map — remap tones to a three-color gradient
GradientOverlay — a two-color gradient wash over your layer
GradientOverlay V2 — the same gradient wash, full 30-mode blend list
InnerGlow — a soft glow radiating in from your subject's edge
InnerGlow V2 — the inward glow effect, full 30-mode blend list
InnerShadow — a Photoshop inset shadow, cast inward
InnerShadow V2 — the inset-shadow effect, full 30-mode blend list
OuterGlow — a soft halo radiating out from your subject
OuterGlow V2 — the outward halo effect, full 30-mode blend list
Stroke — a Photoshop outline around your subject, in one node
Stroke V2 — a Photoshop-style outline, now with the full blend list
AnyRerouter — a reroute that actually keeps the wildcard type
Batch Selector — pull specific images out of a batch by index
Boolean — a true/false value you can wire and display
Boolean Operator (Layer Style)
Boolean Operator V2 — compare two values without a scripting node
Check Mask — did the mask actually find anything?
Check Mask V2 — a finer-grained gate for empty masks
Choice Text Preset (Layer Style)
Color Image (Layer Style)
ColorImage V2 — generate a solid-color image at any size
Color Name — turn a hex color into a word
ColorPicker — pick a color visually, output the hex string
CropBoxResolve — unpack a crop box into plain numbers
CropByMask — zoom into the subject, work at full detail, paste back
CropByMask V2
CropByMask V3
ExtendCanvas — pad an image outward and get an outpaint-ready mask
Extend Canvas V2 (Layer Style)
Float — a numeric value you can wire and display
FluxKontextImageScale — snap an image to Flux Kontext's happy resolutions
GetImageSize — read an image's width and height into your graph
Get Main Colors — pull a 5-color palette out of any image
Get Main Colors V2 — pull the five dominant colors out of any image
GradientImage — generate a two-color gradient from nothing
GradientImage V2 — the same gradient generator, with resolution presets
Gray Value — turn a color into a plain brightness number
H/L Frequency Detail Restore
HSV Value — break a color down into hue, saturation, and value
IC Mask — stitch a reference image into one canvas for in-context editing
IC Mask Crop Back — pull the edited region back out of an IC Mask canvas
If — pick one of two values based on a condition
Image Batch To List(Multi) — split a batch so you can process images one at a time
ImageBlend — the plain layer-over-layer composite node
ImageBlendAdvance — composite a layer with full position, scale and rotate control
ImageBlendAdvance V2 — composite with position, scale, rotate and 30 blend modes
ImageBlendAdvance V3
ImageBlend V2
ImageChannelMerge — rebuild an image from separate color channels
ImageChannelSplit — pull an image apart into its individual color channels
Image Combine Alpha (Layer Style)
Image Composite Handle Mask
ImageHub — a 9-way image switch for tidying up busy graphs
Image List To Batch (Layer Style)
Image Mask Scale As
Image Mask Scale As V2
Image Opacity (Layer Style)
ImageReel — bundle up to four labeled images into a comparison strip
ImageReelComposit — turn your reels into one labeled contact sheet
ImageRemoveAlpha — flattening a transparent image back to RGB
ImageScaleByAspectRatio — aspect-ratio resize, the original
ImageScaleByAspectRatio V2 — resize to any ratio, snapped to sampler-friendly dimensions
Image Scale Restore (Layer Style)
Image Scale Restore V2 (Layer Style)
ImageShift — nudge or wrap-scroll an image, with a border mask thrown in
Image Tagger Save
Image Tagger Save V2 — save an image with its caption as a matching text file
Integer — a plain integer, exposed as both a number and a string
LayerImageTransform — move, scale, rotate and mirror an image on its own canvas
LayerMaskTransform — move, scale, and rotate a mask independent of its image
Load Images From Path
Load VQA Model — load a BLIP model that can answer questions about an image
Name To Color — turn a color name like \"cornflower blue\" into a hex value
Nano Banana Image Scale — resize an image to what Nano Banana actually wants
Number Calculator (Layer Style)
Number Calculator V2
PrintInfo — a debug probe you plug into anything to see what's actually there
PurgeVRAM — dump the model cache between heavy stages
PurgeVRAM V2 — free VRAM mid-workflow so the next stage doesn't OOM
QueueStop — halt the queue mid-graph
Random Generator — one node, three flavors of random
Random Generator V2 — the reproducible version
RestoreCropBox — paste a processed crop back where it came from
RGB Value — split a color into its R, G, B numbers
RoundedRectangle — build a rounded-corner mask, optionally around an object
Seed — one seed value, wired everywhere you need it
SimpleTextImage — render text straight onto an image, no editor needed
String — a plain text constant you can wire anywhere
String Condition (Layer Style)
Switch Case
Text Box (Layer Style)
Text Image (Layer Style)
TextImage V2 — the full-featured text renderer, vertical layout included
TextJoin — glue up to four strings together, no delimiter
TextJoinV2 — glue up to four strings together with a delimiter
Text Preseter — save a title-and-content pair as a reusable preset
VQA Prompt — ask a model specific questions about an image
XY to Percent — turn a pixel position into a resolution-proof percentage
A set of nodes for ComfyUI it generate image like Adobe Photoshop's Layer Style. the Drop Shadow is first completed node, and follow-up work is in progress.