ComfyUI-SBTools

Latest Version: 1.6.0

Custom node collection for ComfyUI. Background removal, color analysis, and dynamic prompt generation tools.

Nodes

| Node | Category | Description | | -------------------------------- | -------------- | ------------------------------------------------------------------------ | | BiRefNet RemoveBG (SBTools) | SBTools/Image | Advanced background removal with 5 model variants | | Alpha to Chroma Key (SBTools) | SBTools/Image | Find safe chroma key colors and fill transparent areas automatically | | Variable Prompt (SBTools) | SBTools/Prompt | Define variables with sequential/random/conditional selection modes | | Variable Folder (SBTools) | SBTools/Image | Define conditional image folder mappings based on variable context | | Variable Image Loader (SBTools) | SBTools/Image | Load images from folder with pattern matching and flexible control | | Variable Combiner (SBTools) | SBTools/Prompt | Combine multiple variable lists for unlimited expansion | | Variable Builder (SBTools) | SBTools/Prompt | Generate prompts and load images with debug info and combination details | | Match Color (SBTools) * | SBTools/Image | Match color distribution between images (Lab/RGB/Adaptive) | | Match Color Balance (SBTools) * | SBTools/Image | Match color temperature and tint (MKL/Shift methods) | | Match Luminance (SBTools) * | SBTools/Image | Match brightness levels between images | | Save Text (SBTools) | SBTools/Text | Save text to file with sequential numbering, overwrite, or append modes |

* In development

Installation

Method 1: ComfyUI Manager (Recommended)

The easiest way to install is via ComfyUI Manager.

1. Open ComfyUI Manager in your ComfyUI interface.
2. Click on "Custom Nodes Manager".
3. Search for "ComfyUI-SBTools".
4. Click "Install".
5. Restart ComfyUI.

Method 2: Manual Installation

Step 1: Open your ComfyUI custom nodes folder

• Windows Portable: ComfyUI_windows_portable\ComfyUI\custom_nodes
• Standard: ComfyUI\custom_nodes

Step 2: Open terminal in this folder and clone the repository

git clone https://github.com/Amatsukast/ComfyUI-SBTools.git

Step 3: Install dependencies

For Windows Portable:

cd ComfyUI-SBTools
..\..\..\python_embeded\python.exe -m pip install -r requirements.txt

For Standard Installation (venv/conda):

cd ComfyUI-SBTools
pip install -r requirements.txt

Step 4: Restart ComfyUI

To update: Navigate to the ComfyUI-SBTools folder and run git pull, then update dependencies with pip install -r requirements.txt and restart ComfyUI.

Variable System

Workflows (Variable System)

Example Workflow 1: Basic Variable Prompt

This example demonstrates the complete prompt generation workflow:

Variables defined:

• GENDER: man, woman (Sequential)
• AGE: young, middle-aged, old (Sequential)
• CLOTHING: suit, lab coat, casual wear (Sequential)
• ACCESSORY: glasses, hat, [NONE] (Random with prefix " and ")

Setup:

• 4 variables combined with Variable Combiner
• Template: "A [AGE] [GENDER] in [CLOTHING][ACCESSORY]."
• Primitive node with increment controls the index for batch processing

Result:

• 18 sequential combinations (2 × 3 × 3 from Sequential variables)
• Random accessory selected each execution (3 choices including [NONE])
• Example output: "A young man in suit and glasses."

Download: Variable Prompt.json

Example Workflow 2: Variable Prompt with Images

This example demonstrates the combined text + image workflow:

Variables defined:

• GENDER: man, woman (Sequential)
• AGE: young, middle-aged, old (Sequential)
• CLOTHING: suit, lab coat, casual wear (Random)
• ACCESSORY: glasses, hat, [NONE] (Random with prefix " and ")
• Image Variable: body reference images from folder (Sequential)

Setup:

• 4 text variables + 1 image variable combined with Variable Combiner
• Variable Builder generates prompts and loads images
• Primitive node with increment controls the index for batch processing

Result:

• Full combinations of text variables × image files
• Example: 2 (gender) × 3 (age) × 3 (body images) = 18 combinations
• Clothing and accessory are randomized for each execution
• Each index outputs corresponding prompt + image

Download: Variable Prompt and Image_1.json

Example Workflow 3: Conditional Variables (Basic)

This example demonstrates the conditional variable system:

Variables defined:

• GENDER: man, woman (Sequential)
• AGE: young, middle-aged, old (Sequential)
• CLOTHING: suit, casual (Sequential, with conditional values)
  • Common: suit, casual
  • [man]: business suit
  • [woman]: dress, skirt
  • [*]: sportswear (back to common)
• ACCESSORY: glasses, watch, [NONE] (Random)

Setup:

• Variables connected sequentially: GENDER → AGE → CLOTHING → ACCESSORY
• Each variable's var_list output connects to the next variable's var_list input
• Variable Builder generates prompts based on conditional logic

Result:

• Clothing options change based on GENDER value
• For "man": suit, casual, business suit, sportswear
• For "woman": suit, casual, dress, skirt, sportswear
• Total combinations automatically calculated

Download: Variable Prompt Conditional_1.json

Example Workflow 4: Conditional Variables (Advanced)

This example demonstrates complex conditional logic with multiple conditions:

Variables defined:

• GENDER: man, woman (Sequential)
• AGE: young, old (Sequential)
• CLOTHING: Multiple conditional sections
  • Common values
  • [man]: men's clothing
  • [woman]: women's clothing
  • [young&&man]: young men's casual wear
  • [old||woman]: elegant clothing for old people OR women
  • [*]: universal clothing items
• ACCESSORY: Context-aware accessories with ConditionalRandom mode

Setup:

• Complex condition syntax: && (AND), || (OR), [*] (wildcard)
• ConditionalRandom mode: Random selection within available conditional values
• Variable Builder shows all combinations with debug output

Result:

• Clothing dynamically filters based on GENDER and AGE
• Example: young man → includes young men's casual wear
• Example: woman (any age) → includes elegant clothing (OR condition matches)
• Accessory randomly selected from context-appropriate options

Download: Variable Prompt Conditional_2.json

Example Workflow 5: Advanced Conditional Syntax

This example demonstrates the three advanced conditional syntax features:

Variables defined:

• GENDER: man, woman (Sequential)
• AGE: young, middle, old (Sequential)
• CLOTHING: Advanced conditional logic
  • Common: suit, t-shirt
  • [man&&!young]: trench coat (NOT condition - excludes young)
  • [man&&young] --only: school uniform (Exclusive mode - ignores common values)
  • [man&&middle]: --t-shirt (Exclusion syntax), jacket

Setup:

• Three advanced syntax features demonstrated:
  1. NOT condition (!): [man&&!young] adds trench coat only for non-young men
  2. Exclusion (--value): --t-shirt removes t-shirt from middle-aged men
  3. Exclusive mode (--only): [man&&young] --only replaces all common values

Result:

• man + young: school uniform only (--only overrides common values)
• man + middle: suit, ~~t-shirt~~ (excluded), jacket, trench coat (NOT young = true)
• man + old: suit, t-shirt, trench coat (NOT young = true)
• woman + *: suit, t-shirt (common values only)

Key Points:

• NOT condition allows "everything except" logic
• Exclusion removes specific items from inherited values
• Exclusive mode creates completely separate value sets for specific conditions

Download: Variable Prompt Conditional_3.json

Example Workflow 6: Conditional Image Loading

This example demonstrates conditional image loading with Variable Folder:

Variables defined:

• GENDER: man, woman (Sequential)
• CLOTHING: Multiple conditional sections
  • Common: suit, casual wear, sportswear
  • [man]: tuxedo
  • [woman]: dress
• ACCESSORY: watch, glasses, [NONE] (Random with prefix "with ")
• Variable Folder (IMAGE1): Context-aware image folder mappings
  • [man]: C:\images\body_man
  • [man&&casual wear]: C:\images\body_man_casual
  • [woman]: C:\images\body_woman

Setup:

• Variable Folder connected after CLOTHING variable
• Variable Image Loader receives folder mappings from Variable Folder
• Variable Builder generates prompts and loads images based on conditions

Result:

• Images automatically switch based on GENDER and CLOTHING context
• Example: "man" + "casual wear" → loads from body_man_casual folder
• Example: "woman" + "suit" → loads from body_woman folder
• Template: "A [GENDER] in [CLOTHING] [ACCESSORY]. Body shape and pose from image 1."

Download: Variable Prompt and Image_2.json

Example Workflow 7: Image Load from Folder

This example demonstrates the Variable Image Loader used as a standalone image loader:

Features:

• Load images from folder with glob pattern matching
• Sequential and Random modes with index/seed control
• Natural sort order (file1, file2, ..., file10)
• Subfolder search with recursive option
• RGBA preservation or background fill with custom color
• Filename output with optional extension toggle

Setup:

• Variable Image Loader with folder path and pattern (e.g., *.png)
• image output connects directly to image processing nodes
• Primitive node with increment controls the index for sequential loading
• total_images and filename outputs for debugging

Result:

• Simple image batch processing without text variables
• Perfect for testing, preprocessing, or sequential image loading
• Can also be used with Variable Combiner for combined workflows

Download: Image Load from Folder.json

Getting Started

System Overview

The prompt and image generation system consists of these nodes:

1. Variable Prompt - Define text variables with their values
2. Variable Folder - Define conditional image folder mappings (optional, for context-aware images)
3. Variable Image Loader - Load images from folder with flexible control
4. Variable Combiner - Combine multiple variables into lists (optional, for complex workflows)
5. Variable Builder - Generate prompts and load images with debug info

Quick Start

Simple Example (2-3 variables):

Variable Prompt (GENDER: man, woman) → Variable Builder
Variable Prompt (AGE: young, old)    ↗

Complex Example (7+ variables):

Variable Prompts 1-3 → Combiner A ┐
Variable Prompts 4-6 → Combiner B ├→ Variable Builder
Variable Prompt 7    ──────────────┘

With conditional images:

Variable Prompt (GENDER) → Variable Folder (Images) → Variable Image Loader → Variable Builder

Conditional Syntax

Variable Prompt and Variable Folder share the same conditional syntax system. This allows dynamic value/folder selection based on previous variable values.

Basic Syntax

| Syntax | Description | Example | Matches When | | -------------------- | --------------------------- | ----------------------------- | --------------------------------------- | | [value] | Simple condition | [man] | GENDER = "man" | | [value1&&value2] | AND condition (both true) | [man&&young] | GENDER = "man" AND AGE = "young" | | [value1\|\|value2] | OR condition (any true) | [woman\|\|old] | GENDER = "woman" OR AGE = "old" | | [value&&!other] | NOT condition (negation) | [man&&!young] | GENDER = "man" AND AGE ≠ "young" | | [TAG:value] | Tag specification | [GENDER:man&&CLOTHING:suit] | Explicit tag names for duplicate values | | [*] | Wildcard (return to common) | [*] | End conditional section |

Syntax Variants:

• AND: &&, AND (uppercase only), ＆＆ (full-width)
• OR: ||, OR (uppercase only), ｜｜ (full-width)
• Tag separator: :, ： (full-width)

Example:

suit
casual

[man]
tuxedo

[woman]
dress

[young&&man]
hoodie

[old||woman]
elegant coat

[*]
sportswear

Result:

• young man: suit, casual, tuxedo, hoodie, sportswear
• old man: suit, casual, tuxedo, elegant coat, sportswear
• young woman: suit, casual, dress, elegant coat, sportswear
• old woman: suit, casual, dress, elegant coat, sportswear

Advanced Syntax

| Syntax | Placement | Description | Example | Result | | --------- | -------------------- | -------------------------------------------- | ----------------------------------------------- | --------------------------------------------------------- | | --only | After condition line | Ignores all common values for this condition | [man&&young] --only<br>school uniform | man+young: "school uniform" only<br>others: common values | | --value | In value list | Removes specific value/folder from context | [man&&middle]<br>--t-shirt<br>formal vest | man+middle: "suit, ~~t-shirt~~, jacket, formal vest" |

Example for --only:

suit
casual

[man&&young] --only
school uniform

Result: man+young gets only "school uniform", others get "suit, casual"

Example for --value:

suit
t-shirt
jacket

[man&&middle]
--t-shirt
formal vest

Result: man+middle gets "suit, ~~t-shirt~~, jacket, formal vest"

Usage Notes

• Case sensitivity: Condition values are case-sensitive
• Connection required: Connect var_list output to enable conditional logic
• Order matters: Use [*] to return to common values after conditional sections
• Combining features: All syntax features can be combined (NOT + --only, etc.)
• Best practices:
  • Use --only for complete replacement (clearer intent)
  • Use --value for removing specific items
  • Use ! for "everything except" logic

Node Reference (Variable System)

Variable Prompt

Define a single variable with multiple values. Variables can operate in four modes: Sequential, Random, Conditional, and ConditionalRandom.

Parameters:

Required:

• tag_name - Variable name for template replacement (e.g., GENDER, CLOTHING)
  • Leave empty for auto-naming (_VAR_xxxxxx) - useful when not using template replacement
  • Auto-named variables are exempt from duplicate checking
• values - List of values, one per line
  • Use empty line or [NONE] for "no value" option
  • Use [condition] syntax for conditional values (see Conditional Syntax section)
• randomize - Toggle between modes:
  • OFF (Sequential/Conditional): Cycle through all values systematically
  • ON (Random/ConditionalRandom): Pick one value randomly each execution

Optional:

• prefix - Text added before the value (only in template mode)
• suffix - Text added after the value (only in template mode)
• var_list - Connect previous variable to enable conditional logic based on its values

Important Notes:

• Tag Name Conflicts: Using the same tag name with different values in Variable Combiner will cause an error. To avoid conflicts, either use unique names or leave tag_name empty for auto-naming.
• Conditional Variable Warnings: If a condition doesn't match any previous variable values, a warning will be displayed.

Output:

• var_list - Variable data (connect to Variable Folder, Variable Combiner, or Variable Builder)

Examples:

Basic variable:

tag_name: "GENDER"
values: "man\nwoman"
randomize: OFF

With prefix/suffix:

tag_name: "ACCESSORY"
values: "glasses\nhat\n[NONE]"
prefix: " wearing "
randomize: ON

Conditional Variables:

Variables can change their available values based on previous variable values. Connect the var_list output from a previous variable to enable conditional logic.

Basic Syntax:

common_value1
common_value2
[condition]
conditional_value1
conditional_value2
[*]
back_to_common

Example:

Variable 1 (GENDER):
  man
  woman

Variable 2 (CLOTHING) - var_list connected to Variable 1:
  suit
  casual
  [woman]
  dress
  [*]
  sportswear

Result:

• If GENDER = "man": Available = suit, casual, sportswear
• If GENDER = "woman": Available = suit, casual, dress, sportswear

With Random Mode:

Set randomize: ON on the conditional variable to enable ConditionalRandom mode. The system will:

1. Determine available values based on current conditions
2. Randomly select one from the available set

For detailed conditional syntax (AND, OR, NOT, --only, exclusion, etc.), see the Conditional Syntax section above.

Variable Folder

Define conditional image folder mappings that change based on variable context. Works like Variable Prompt but for image folders — different conditions load images from different folders.

Parameters:

Required:

• variable_name - Variable name for the image (e.g., IMAGE1, BODY_IMAGE)
  • Leave empty for auto-naming (_IMAGE_xxxxxx)
• folder_map - Conditional folder definitions with syntax (see below)

Optional:

• var_list - Connect previous variable to enable conditional logic based on its values

Syntax:

Basic Structure:

[condition] --mode_options
folder_path1 --path_options
folder_path2 --path_options

[another_condition]
folder_path3

For conditional syntax details (AND, OR, NOT, --only, etc.), see the Conditional Syntax section above.

Folder-Specific Options:

Condition Options (placed after ]):

• --random - Random selection within this context
• --sequential - Sequential selection (default)

Path Options (placed after folder path):

• --subfolder - Include subfolders in search
• --pattern=*.png - File pattern (e.g., *.png, body_*.jpg)
• --extension - Include file extension in filename output
• --no-extension - Exclude extension (default)
• --fill-bg=#FFFFFF - Fill transparent areas with specified color
• --no-fill-bg - Keep transparency (default)

Basic Example:

[man]
C:\images\body_man

[man&&casual wear]
C:\images\body_man_casual --subfolder

[woman]
C:\images\body_woman --pattern=*.png --fill-bg=#FFFFFF

[*]
C:\images\body_common

Workflow Connection:

Important: Variable Folder must be connected to Variable Image Loader to work.

Variable Prompt (GENDER) ┐
Variable Prompt (CLOTHING) ├→ Variable Folder → Variable Image Loader → Variable Builder

Connection Behavior:

• Variable Folder connected: Folder mappings from Variable Folder are always used, regardless of Variable Image Loader's folder_path setting
• Variable Folder disconnected or bypassed: Variable Image Loader's own folder_path parameter becomes active
• No matching images: If a condition has no matching images, that combination outputs an empty image (64×64 black image)

Tips:

Setting up conditions:

• Conditions must match values from previous variables (case-sensitive)
• Use var_list connection to enable conditional logic
• Warnings appear if conditions don't match any previous values

Organizing folders:

• Use specific conditions first, then general ones
• Example: [man&&casual] before [man]
• Use [*] to return to common folders after conditional sections

Path configuration:

• Absolute paths recommended for clarity
• Relative paths work from ComfyUI root
• Use --subfolder with --pattern=**/*.png for recursive search

For advanced conditional syntax (NOT, --only, exclusion), see the Conditional Syntax section above.

Variable Image Loader

Load images from a folder with flexible control. Can be used standalone or combined with text variables for full workflow generation. When connected with Variable Folder, automatically switches to conditional image loading mode.

Parameters:

Required:

• folder_path - Path to folder containing images (absolute or relative)
  • Note: When Variable Folder is connected, this parameter is ignored and folder mappings from Variable Folder are used instead
• pattern - File pattern for glob matching (e.g., *.png, body_*.jpg)
• index - Index to select which image in sequential mode (loops automatically)
• randomize - Toggle between modes:
  • OFF (Sequential): Cycle through images by index
  • ON (Random): Pick one image randomly based on seed
• seed - Seed for random selection

Optional:

• var_list - Connect from Variable Folder to enable conditional image loading
• include_subfolders - Search in subfolders (requires **/ in pattern, default: OFF)
• include_extension - Include file extension in filename output (default: OFF)
• fill_background - Fill transparent areas with solid color (default: OFF, keeps RGBA)
• background_color - Background color in hex format (e.g., #FFFFFF for white)

Outputs:

• image - Loaded image (IMAGE type, for standalone use)
• var_list - Variable list (for Variable Combiner)
• total_images - Total number of images found
• filename - Current filename (with or without extension)

Features:

• Natural sort order: Files are sorted like Windows Explorer (e.g., file1, file2, ..., file10)
• RGBA preservation: Transparent images keep transparency by default
• Multibyte support: Japanese and other multibyte characters in paths/filenames work correctly
• Dual output: Use as standalone image loader OR as variable for combination workflows
• Variable Folder integration: When Variable Folder is connected, automatically applies conditional folder mappings

Connection Modes:

Standalone Mode (no var_list connected):

• Uses folder_path parameter to load images
• Simple folder-based image loading

Variable Folder Mode (var_list connected from Variable Folder):

• Folder mappings from Variable Folder are always used
• folder_path parameter is ignored
• Images automatically switch based on variable context
• Loader settings (randomize, seed, fill_background, etc.) are applied to Variable Folder configuration

Examples:

Standalone image loader:

Variable Image Loader (folder: body_refs/, pattern: *.png)
  ├─ image → FLUX2
  └─ index ← Primitive (increment)

Combined with text variables:

Variable Prompt (GENDER) ┐
Variable Prompt (AGE)    ├→ Variable Combiner → Variable Builder
Variable Image Loader    ┘                            ↓
                                              prompt + image

Conditional image loading with Variable Folder:

Variable Prompt (GENDER) → Variable Folder → Variable Image Loader → Variable Builder
                                                     ↓
                                            Context-aware images

Variable Combiner

Combine multiple variable lists into one. Useful for organizing complex prompts with many variables.

Parameters:

Optional:

• var_list1 to var_list6 - Variable lists from Variable Prompt, Variable Folder, Variable Image Loader, or other Combiners

Output:

• var_list - Combined variable list

Usage:

Organize by category:

Character variables (3) → Combiner A ┐
Clothing variables (3)  → Combiner B ├→ Combiner C → Variable Builder
Scene variables (2)     → Combiner C ┘

Chain for unlimited expansion:

• Each Combiner supports 6 inputs
• Connect Combiner outputs to other Combiners
• No limit on total number of variables

Duplicate Detection:

Variable Combiner automatically checks for duplicate variables:

Exact duplicates (same name, values, and mode):

• Automatically skipped with info message
• Second occurrence is ignored
• Example: Same GENDER variable used in multiple branches

Conflicting duplicates (same name, different values):

• Throws error and stops execution
• Must be fixed by renaming or using empty tag names
• Example: Two different definitions of GENDER

Auto-named variables (_VAR_*):

• Always exempt from duplicate checking
• Can appear multiple times without conflict

Variable Builder

Generate prompts and load images with full debug information. Supports text-only workflows, image workflows, or combined workflows.

Parameters:

Required:

• template - Template text with [TAG_NAME] placeholders
  • Leave empty for simple join mode
  • Example: "A [AGE] [GENDER] wearing [CLOTHING]."
• index - Index to select which Sequential combination (loops automatically)
• seed - Seed for Random TEXT variables (Image Variables have independent seeds)
• separator - Character(s) to join values (default: ", ")

Optional:

• var_list - Variable list from Variable Prompt, Variable Folder, Variable Image Loader, or Variable Combiner

Outputs:

• prompt - Generated prompt text (STRING)
• image1-4 - Loaded images (IMAGE × 4, empty if no image variables)
• max_combinations - Total number of sequential combinations (INT)
• all_combinations - Debug text listing all patterns with index numbers (STRING)

Modes:

Template Mode (template not empty):

• Tags like [TAG_NAME] are replaced with values
• Unused variables are appended at the end

Simple Join Mode (template empty):

• All values are joined with separator

Features:

• Text-only or image workflows: Works with any combination of text and image variables
• Up to 4 images: Perfect for FLUX.2 Reference workflow with multiple reference images
• Independent seed control: Each Image Variable uses its own seed, text variables use Variable Builder's seed
• Full combination calculation: All Sequential text × Sequential images × Random variations
• Debug output: max_combinations shows total patterns, all_combinations lists all with index
• Random preview: Shows [RANDOM: choice1|choice2|...] for random variables in debug output
• Empty slot handling: Unused image slots automatically filled with blank images

Example Workflow:

For FLUX.2 Reference with multiple images:

Variable Prompt (STYLE)           ┐
Variable Prompt (POSE)            ├→ Variable Combiner
Variable Image Loader (Body)      ┤       ↓
Variable Image Loader (Face)      ┘  Variable Builder
                                          ↓
                              prompt + image1 + image2
                                          ↓
                                     FLUX2 with
                                  Reference Latent

With conditional images:

Variable Prompt (GENDER) → Variable Folder (Body) → Variable Image Loader ┐
Variable Prompt (POSE)                                                    ├→ Variable Combiner → Variable Builder
Variable Image Loader (Face)                                              ┘

Combination calculation:

STYLE: 3 values (Sequential)
POSE: 4 values (Sequential)
Body images: 5 files (Sequential)
Face images: 3 files (Sequential)

Total: 3 × 4 × 5 × 3 = 180 combinations

Tips:

Batch processing:

• Connect Primitive (INT, increment) to index for sequential patterns
• Connect Primitive (INT, increment) to seed for random variations
• Use max_combinations output to know total patterns
• Connect all_combinations to Show Text node to see all patterns

For FLUX.2 JSON-style prompts:

{
  "subject": "[SUBJECT]",
  "background": "[BACKGROUND]",
  "lighting": "[LIGHTING]",
  "style": "[STYLE]"
}

For natural language:

"A [AGE] [GENDER] [CLOTHING][ACCESSORY], [BACKGROUND], [LIGHTING]"

Notes:

• Image Variables maintain their own randomize/seed settings
• Empty image slots output blank images when using fewer than 4 Image Variables
• Use same index from Primitive (increment) for synchronized batch processing
• Performance: 10,000+ combinations enumerate quickly (under 1 second)

Image Processing

Workflows (Image Processing)

Example Workflow: Background Removal and Chroma Key Fill

This example demonstrates the complete background removal and chroma key workflow:

Workflow:

1. BiRefNet RemoveBG - Removes background and creates transparent PNG
2. Alpha to Chroma Key - Finds unused color safe for chroma keying
3. Filled image output - Automatically fills transparent areas with the detected color

Features:

• Automatic safe color detection that doesn't conflict with foreground
• One-click background fill for chroma key compositing
• Perfect for video editing software that requires solid backgrounds
• Color visualization output shows the detected color

Use Case:

• Remove background from photos/renders
• Prepare images for video editing (After Effects, Premiere, DaVinci Resolve)
• Create green screen / blue screen equivalents with optimal colors
• Avoid color conflicts with subject colors

Download: Remove BG and Fill BG.json

Node Reference (Image Processing)

BiRefNet (RemoveBG)

Advanced background removal using BiRefNet models. Supports multiple model variants optimized for different use cases.

Available Models:

| Model | Best For | Resolution | Notes | | --------------------- | ------------------------- | --------------------- | --------------------------- | | BiRefNet-general | Everyday use | 1024×1024 | Fast, balanced performance | | BiRefNet-HR | High quality | 2048×2048 | Best detail preservation | | BiRefNet-portrait | People/portraits | 1024×1024 | Trained on human subjects | | BiRefNet_dynamic | Aspect ratio preservation | Variable (256-2304px) | No image distortion | | BiRefNet_toonout | Outline extraction | 1024×1024 | Creates toon-style outlines |

Parameters:

Required:

• image - Input image
• model - Model selection

Optional:

• mask_blur (0-64) - Softens mask edges. Use for smoother transitions.
• mask_offset (-20 to +20) - Expands (+) or shrinks (-) the mask boundary
• invert_output - Swaps foreground/background
• background - Alpha (transparent) or Color (custom color)
• background_color - Custom background color (hex code, e.g., #222222)

Outputs:

• IMAGE - Processed image with transparent or colored background
• MASK - Black & white mask data
• MASK_IMAGE - Visualization of the mask (RGB)

Which Model Should I Use?

For most cases: Use BiRefNet-general

• Fast processing
• Good quality
• Works well for general objects

For high-quality results: Use BiRefNet-HR

• 4x larger processing area
• Better detail preservation
• Ideal for final outputs

For people/portraits: Use BiRefNet-portrait

• Optimized for human subjects
• Better hair/skin detection
• Trained on portrait datasets

For wide/tall images: Use BiRefNet_dynamic

• Preserves aspect ratio (no squashing/stretching)
• Automatically scales to fit
• Best for non-square images

For creative effects: Use BiRefNet_toonout

• Extracts outlines instead of removing background
• Creates line art/edge detection
• Useful for stylized effects

Tips:

Processing Resolution:

• Square models (general, HR, portrait, toonout) resize images to a fixed square
• Dynamic model preserves aspect ratio by resizing the longest side only

Mask Refinement:

• Start with mask_blur: 0 and mask_offset: 0
• Use mask_blur: 2-5 for softer, more natural edges
• Use mask_offset: +2 to +5 if edges are too tight
• Use mask_offset: -2 to -5 if edges include too much background

Performance:

• Smaller images process faster
• BiRefNet-general is fastest
• BiRefNet-HR takes ~4x longer but produces better results

Alpha to Chroma Key

Finds a color in your image that is maximally different from all existing colors, and automatically fills transparent areas with that color. Perfect for chroma keying workflows in video editing.

Use Cases:

• Chroma key backgrounds - Find a color that won't interfere with your subject and fill transparent areas
• Video editing preparation - Create solid backgrounds for After Effects, Premiere, DaVinci Resolve
• Green screen replacement - Generate optimal chroma key colors automatically
• Mask generation - Create temporary backgrounds for selection tools

Parameters:

Required:

• image - Input image to analyze (typically with transparent background)

Optional:

• min_distance (0-255, default: 30) - Minimum color distance required
  • Higher values = more different from existing colors
  • Lower values = easier to find but may be similar to image colors
• sample_size (1000-50000, default: 5000) - Number of pixels to sample
  • Higher values = more accurate but slower
  • Lower values = faster but may miss similar colors

Outputs:

• hex_color - Color in hex format (e.g., #00FF00)
• filled_image - Image with transparent areas filled with the detected color
  • Perfect for chroma key workflows - transparent areas are automatically filled
  • Output is ready for video editing software

How It Works:

1. Samples random pixels from your image
2. Tests common pure colors (green, blue, magenta, cyan, yellow, red)
3. Returns the color with the maximum distance from all sampled pixels
4. Automatically fills transparent areas (alpha channel) with the detected color
5. If no candidate meets the criteria, performs a coarse grid search

Typical Workflow:

Load Image → BiRefNet RemoveBG → Alpha to Chroma Key
                                          ↓
                                  filled_image (ready for video editing)
                                  hex_color (for reference)

See the complete workflow in the Example Workflow section above.

Tips:

For chroma keying:

• Use min_distance: 40-60 for safe separation from subject colors
• Pure green (0, 255, 0) is usually selected for typical images
• The algorithm automatically avoids colors present in your subject

For quick results:

• Use sample_size: 5000 (default) for most images
• Increase to 10000-20000 for images with complex color palettes

For video editing:

• Use filled_image output directly in your video editor
• Reference hex_color output if you need to adjust keying settings
• Works with any video editing software that supports chroma keying

Distance metric:

• Distance is calculated as Euclidean distance in RGB space
• A distance of 30 means the color differs by ~30 units per channel on average

Utilities

Save Text

Save any text string from your workflow to a local file. Designed as a terminal node — connect a STRING output from another node and save it with full control over path, filename, and write behavior.

Parameters:

Required:

• text - Text to save (connect from another node)
• folder_path - Subfolder relative to ComfyUI output directory (supports strftime, e.g. %Y/%m/%d)
• filename_prefix - Base filename (supports strftime, e.g. %Y%m%d_%H%M%S). Default: text_output
• extension - File extension without dot. Default: txt
• save_mode - Write behavior:
  • Sequential: Append zero-padded counter to filename; always creates a new file
  • Overwrite: Replace existing file with the same name
  • Append: Add text to end of existing file (inserts newline only when file already has content)
• counter_digits - Digits for zero-padding (1–9). Only used in Sequential mode. Default: 3
• counter_position - Counter placement: Back → name_001.txt, Front → 001_name.txt. Only used in Sequential mode
• separator - Character between filename and counter. Default: _

Output: None (terminal node)

Notes:

• folder_path and filename_prefix both support Python strftime formatting
• Path traversal (../) is blocked; all files are saved inside the ComfyUI output directory
• counter_digits, counter_position, and separator are only active in Sequential mode
• Files are always saved with UTF-8 encoding

Download: Save Text.json

Technical Details

Model Storage

BiRefNet models are automatically downloaded and stored in:

ComfyUI/models/sbtools/BiRefNet/

Models are downloaded on first use and cached for future sessions.

Credits & License

Credits

BiRefNet Node is based on ComfyUI-RMBG by AILab-AI

BiRefNet Models by ZhengPeng7 - HuggingFace | GitHub

License

Code License

This project is licensed under GNU General Public License v3.0 (GPL-3.0).

• BiRefNet node is based on ComfyUI-RMBG by AILab-AI (GPL-3.0)
• Find Unused Color node is original work by Amatsukast

See LICENSE for details.

Model License

BiRefNet models by ZhengPeng7 are licensed under Apache License 2.0.

• Model repository: ZhengPeng7/BiRefNet
• You can use the models freely for commercial and non-commercial purposes

Summary

• ✅ You can use, modify, and distribute this code
• ✅ You can use the models commercially
• ⚠️ Modifications must also be GPL-3.0
• ⚠️ You must provide source code when distributing

Changelog

See CHANGELOG.md for detailed version history.

Version History

v1.6.0 (2026-05-27) - Utilities

• Save Text node for saving workflow text output to file
• Sequential numbering, Overwrite, and Append write modes
• strftime format support for folder and filename
• Path traversal protection

v1.5.1 (2026-04-27) - Advanced Conditional Syntax

• NOT condition (!) for negation logic
• Exclusion syntax (--value) to remove specific values
• Exclusive mode (--only) to override common values
• All three features work in both Variable Prompt and Variable Folder
• New example workflow: Variable Prompt Conditional_3

v1.5.0 (2026-04-24) - Conditional Image Folders

• Variable Folder node for context-aware image loading
• Enhanced combination calculation algorithm
• Variable order control

v1.4.1 (2026-04-21) - Code Refactoring

• File/class name alignment with node IDs
• Improved maintainability

v1.4.0 (2026-04-20) - Conditional Variables

• Conditional variable system with AND/OR logic
• Variable Builder with debug output
• ConditionalRandom mode

v1.3.0 (2026-04-17) - Image Variables

• Variable Image Loader node
• Variable Builder for text + image workflows
• Support for up to 4 images

v1.2.0 (2026-04-17) - Variable System

• Variable Prompt, Variable Combiner nodes
• Template-based prompt generation

v1.1.0 (2026-04-13) - Node Updates

• Unique node names with SBTools prefix
• Alpha to Chroma Key filled_image output

v1.0.0 (2026-04-13) - Initial Release

• BiRefNet background removal (5 models)
• Alpha to Chroma Key node

Example workflows included in examples/ folder:

Prompt Generation:

• Variable Prompt.json - Basic text-only prompt generation
• Variable Prompt and Image_1.json - Combined text + image workflow
• Variable Prompt and Image_2.json - Conditional image loading with Variable Folder
• Variable Prompt Conditional_1.json - Basic conditional variables
• Variable Prompt Conditional_2.json - Advanced conditional logic (AND, OR, wildcard)
• Variable Prompt Conditional_3.json - Advanced conditional syntax (NOT, Exclusion, --only)
• Image Load from Folder.json - Standalone image loader

Image Processing:

• Remove BG and Fill BG.json - Background removal with chroma key fill

Utilities:

• Save Text.json - Save text output to file

---

Note: This is a personal custom node collection. Use at your own discretion.