ComfyUI-nhknodes

AddHeadroom

Reduces image size by a percentage while keeping the original canvas dimensions. Image is anchored bottom-center with selectable background fill (white/black/transparent).

Nhk/Image · 1 input · 2 parameters · 1 output

Bookmark

Canvas bookmark with keyboard shortcut navigation (double-click or press shortcut to jump)

Nhk/Utility · 0 inputs · 3 parameters · 0 outputs

ConditionalRouter

Selects between two inputs based on boolean condition with lazy evaluation

Nhk/Wip

ConditionalRouterDual

Selects between two input pairs based on boolean condition with lazy evaluation

Nhk/Wip

ConditionalSplitter

Splits input to two outputs based on condition, blocking the inactive one

Nhk/Wip

ConditionalStop

Silently stops workflow execution based on condition

Nhk/Wip

CropToMask

Crop image+mask to mask bbox with uniform padding (great after SAM3)

Nhk/Image · 2 inputs · 3 parameters · 6 outputs

CyclingSwitch

Automatically cycles through unlimited connected inputs with configurable stay duration

Nhk/Utility · 0 inputs · 2 parameters

DoubleSwitch

Universal switch with two pairs of inputs (A1/A2 and B1/B2)

Nhk/Utility

DoubleSwitchOut

Takes one double input and switches between two different double outputs

Nhk/Utility

EditWithKrita

Sends input images to Krita for editing, then reloads and returns them; when editing is disabled, passes the input through unchanged.

Nhk/Image · 1 input · 4 parameters · 1 output

ExecutionCounter

Counts workflow executions and stops queue when target is reached

Nhk/Utility

ExtractGridPanel

Extracts a single panel from a grid image. Panel notation: letter + number (e.g., a1, b2, c3) - Letter = column (a=first column, b=second column, etc.) - Number = row (1=first row, 2=second row, etc.)

Nhk/Image · 1 input · 4 parameters · 4 outputs

GeminiImageChat

Generate images or analyze images with Gemini

Nhk/Ai · 4 inputs · 8 parameters · 2 outputs

ImageEvaluator

AI-powered image evaluation with confidence scoring for workflow control

Nhk/Wip · 1 input · 4 parameters · 5 outputs

ImageGridBatch

Combines unlimited images into a single batch tensor with dynamic inputs

Nhk/Image · 0 inputs · 0 parameters · 2 outputs

ImageGridComposite

Creates visual grid layouts from unlimited images with dynamic inputs

Nhk/Image · 0 inputs · 4 parameters · 3 outputs

ImageLoaderWithPreviews

Load an image with preview functionality from any folder. Select folder path and get image previews on hover.

Nhk/Image · 0 inputs · 3 parameters · 3 outputs

IntervalGate

Controls workflow branching with interval-based on/off switching

Nhk/Utility

ListSelector

Select from list by index, use default, or batch entire list

Nhk/Utility · 1 input · 3 parameters · 1 output

LLMChat

Chat with OpenAI GPT-4o with vision capabilities

Nhk/Ai · 1 input · 3 parameters · 1 output

Load2x2GridSeries

Loads images from a directory, treating each as a 2x2 grid. Extracts grid cells in order: upper-left, upper-right, bottom-left, bottom-right. Two modes: - single_image: Load specific grid cell by index (use control_after_generate=increment to auto-advance) - random: Load random grid cell using seed Features: - reset: Set to True to reset counter to first grid cell - Outputs current_index and total_images (total grid cells across all files) - label: Unique identifier for tracking separate sequences independently

Nhk/Image · 0 inputs · 8 parameters · 6 outputs

Load3x3GridSeries

Loads images from a directory, treating each as a 3x3 grid. Extracts grid cells in reading order (left-to-right, top-to-bottom). Two modes: - single_image: Load specific grid cell by index (use control_after_generate=increment to auto-advance) - random: Load random grid cell using seed Features: - reset: Set to True to reset counter to first grid cell - Outputs current_index and total_images (total grid cells across all files) - label: Unique identifier for tracking separate sequences independently

Nhk/Image · 0 inputs · 8 parameters · 6 outputs

LoadImageSeries

Loads images from a directory with two modes: - single_image: Load specific image by index (use control_after_generate=increment to auto-advance) - random: Load random image using seed Features: - reset: Set to True to reset counter to first image - Outputs current_index and total_images for progress tracking - label: Unique identifier for tracking separate sequences independently

Nhk/Image · 0 inputs · 7 parameters · 4 outputs

LoadImageSeriesWithPrompts

Loads images from a directory with two modes (single_image or random) and attaches prompts from a markdown lookup file. Features: - reset: Set to True to reset counter to first image - label: Unique identifier for tracking separate sequences independently - default_prompt input is used when no file-specific prompt exists

Nhk/Image · 0 inputs · 9 parameters · 5 outputs

LoadTextFiles

Load text files with auto-incrementing index and folder scanning

Nhk/Text · 0 inputs · 8 parameters · 5 outputs

OllamaChat

Chat with local Ollama models (text and/or vision)

Nhk/Ai · 3 outputs

PlaySound

Passes its input list through to its output and exposes mode, volume, and sound-file inputs.

Nhk/Utility

PreviewImagePassthrough

Preview images and pass them through for conditional routing

Nhk/Wip · 1 input · 0 parameters · 1 output

RGBAFlatten

Flatten RGBA to RGB over solid background, eliminating alpha artifacts

Nhk/Image · 2 inputs · 3 parameters · 1 output

SaveImagePassthrough

Save images and pass them through for conditional routing

Nhk/Wip · 1 input · 1 parameter · 2 outputs

SimpleTextInput

Simple text input field that passes through text without any processing

Nhk/Text · 0 inputs · 1 parameter · 1 output

SizePicker

Comprehensive size picker with model-optimized presets for Flux, Qwen, and SDXL

Nhk/Utility · 0 inputs · 4 parameters · 4 outputs

TextCombiner

Combines unlimited text inputs into a single output with dynamic inputs

Nhk/Text · 0 inputs · 1 parameter · 1 output

TextDisplay

Interactive text viewer with UI display and pass-through functionality

Nhk/Text · 1 input · 0 parameters · 1 output

TextEncodeFlux2ImageEdit

Encodes prompt and reference images for Flux2 Klein image editing. Returns both positive (with prompt) and negative (empty) conditioning with reference latents attached.

Nhk/Conditioning · 5 inputs · 1 parameter · 2 outputs

TextTemplate

Text template engine with placeholder replacement using bracket syntax

Nhk/Text · 4 inputs · 1 parameter · 1 output

TextTemplateExtended

Extended text template engine with 8 placeholder inputs

Nhk/Text · 8 inputs · 1 parameter · 1 output

VisualResizer

Resizes and positions an image within a custom canvas. Allows scaling the input image and positioning it anywhere on the canvas.

Nhk/Image · 1 input · 6 parameters · 1 output