ComfyUI-NN-custom-nodes

NN-custom-nodes

Execution Profiler

A lightweight execution profiler for ComfyUI that tracks node processing times and displays performance information directly in the workflow editor.

• Measures execution time for every executed node
• Displays timing badges on nodes in the frontend
• Tracks nested subgraph execution time
• Logs slow nodes to the console
• Configurable timing thresholds and colors
• Sends real-time profiler events to the UI

During execution, nodes receive live timing information:

• Node execution duration
• Slow node highlighting
• Subgraph accumulated execution time
• Execution status events

This helps identify workflow bottlenecks and optimize slow pipelines.

---

📐 Image Resolution

Calculates an image resolution based on width, height, aspect ratio, and orientation settings.

• Supports common aspect ratios
• Automatically calculates missing dimensions
• Maintains target pixel count when changing aspect ratios
• Rounds dimensions to a configurable divisor
• Supports portrait and landscape orientation

---

🎬 Video Resolution

Calculates video dimensions and frame count based on resolution, aspect ratio, duration, and frame rate.

• Supports image/video aspect ratios
• Calculates total frames automatically
• Keeps dimensions aligned for video models
• Supports portrait and landscape videos

---

🔍 Resolution Scale

Scales an existing resolution while keeping dimensions aligned and optionally limiting maximum size.

• Upscale or downscale resolutions
• Maintains aspect ratio
• Prevents exceeding maximum resolution
• Rounds dimensions for model compatibility

---

🎲 Wildcard Processor

A dynamic prompt expansion system for creating randomized prompts using wildcard files, selection blocks, variables, and reusable prompt templates.

Useful for:

• Prompt variation generation
• Batch image generation
• Large prompt libraries
• Reproducible randomized workflows

Features:

• Random choice blocks

  • Select random options using {one|two|three}
  • Select multiple options with count syntax {2$$one|two|three} or {1-2$$one|two|three}
  • Support variable-length selections with custom separator {1-2$$, $$one|two|three|four}

• Weighted prompt options

  • Control selection probability using weighted choices:
    • {0.8::red|0.2::blue}

• External wildcard files

  • Load reusable prompt lists from wildcard text files
  • Supports nested wildcard paths:
    • __characters/age__
    • __styles/*__

• Custom variables

  • Define reusable prompt values:
    • {!style=cinematic portrait!}
  • Reuse variables:
    • {$style}

• Seed-based reproducible results

  • Same seed produces the same wildcard selections and prompt expansion

• LoRA tag cleanup

  • Automatically removes LoRA tags:
    • <lora:name:strength>
• • Comment handling
  • Full-line comments using #
  • Hash symbols inside prompts are preserved:
    • # character ... → ignored
    • dark #1 style → kept
  • Inline block comments:
    • portrait /* test variation */

---

✨ Image CASharpening (Batch)

Applies Contrast Adaptive Sharpening (CAS) to images with batch processing support.

• GPU accelerated sharpening
• Batch-friendly processing
• Adjustable sharpening strength
• Optional CPU/GPU execution

---

✂️ Image Inset Crop

Crops an image by percentage values from each side.

• Percentage-based cropping
• Supports image batches
• Keeps crop positioning predictable

---

🔍 Image Upscale By

Upscales images using high-quality bicubic interpolation with anti-aliasing.

• Supports large batches
• GPU/CPU processing selection
• Memory-efficient chunk processing

---

👁️ Image Preview Gate

Provides a fallback preview image when no image is connected.

Useful for optional image inputs in workflows.

---

🎨 Image Color Range

Expands or remaps image color range.

Useful for restoring contrast from limited-range images.

---

🎚️ Image Adjust Contrast/Brightness

Adjusts image brightness and contrast.

---

🌈 Image Adjust Color

Adjusts individual RGB channel intensity.

---

🎛️ Image Adjust

Combined image adjustment node.

Supports:

• Brightness
• Contrast
• Saturation
• RGB channel adjustments

---

🎯 Contrast Adjust (Ref)

Matches image contrast to a reference image using luminance statistics.

Useful for:

• Style matching
• Consistent image batches
• Reference-based workflows

---

🖼️ Image Grid

Creates a grid layout from multiple images.

• Supports image batches
• Automatic cropping to target aspect ratio
• Configurable columns
• Resolution limiting

---

🎞️ Video Timeline

Creates a contact sheet/timeline preview from video frames.

Useful for:

• Reviewing generated animations
• Comparing frame sequences
• Debugging video workflows

---

📐 Image Resize (Batch)

Resizes image batches with resolution limits and model-friendly dimension alignment.

• Batch processing
• GPU acceleration
• Automatic maximum resolution limiting
• Dimension rounding support

---

💾 Image Saver

An advanced image saving node with support for multiple formats, automatic filename generation, batch saving, and workflow metadata preservation.

• Supports PNG, JPEG, and WebP output formats
• Automatic timestamp-based filenames
• Handles image batches
• Prevents filename collisions by creating unique filenames
• Preserves ComfyUI workflow and prompt metadata
• Supports PNG transparency (RGBA)
• Configurable quality settings
• Optional lossless WebP output

Supports dynamic filename placeholders:

• %date

  • Inserts the current date
  • Format: YYYY_MM_DD
  • Example: 2026_08_07

• %timestamp

  • Inserts the current date and time
  • Format: YYYY-MM-DD_HH-MM-SS
  • Example: 2026-08-07_00-47-30

• %time

  • Inserts the current time
  • Format: HH-MM-SS
  • Example: 00-47-30

---

📁 Path Info

Extracts information from a file path and separates it into directory, filename, and path components.

Useful for workflows that need to manipulate saved image paths or process generated files.

---

🧹 Soft Clear Cache

Releases unused ComfyUI resources by unloading models, clearing GPU cache, and running garbage collection. Useful for long workflows or when managing limited VRAM.

---

🖼️ Image Optimize

Optimizes image tensors by ensuring they are stored in contiguous memory format. This can improve compatibility and performance for some operations.

---

🔍 InputCheck

Checks whether an input value is connected or empty.

---

🔀 Bool Switch

Selects between two inputs based on a boolean condition.

Uses lazy evaluation, meaning only the selected input is evaluated when needed.

---

Configuration

The extension uses a settings.toml file to configure global behavior for wildcards and the execution profiler.

Configuration file:

---

Wildcards Settings

[wildcards]

# Directory containing wildcard text files
path = "wildcards"

# Cache loaded wildcard files
# Disable to reload wildcard files without restarting ComfyUI
cache = true

# Remove LoRA tags from processed prompts
cleanup = true

# Separator used when selecting multiple wildcard options
delimiter = ", "

# Maximum recursive expansion passes
max_iterations = 5

Profiler Settings

[profiler]

# Enable execution profiler
enable = true

# Print execution times to console
console_log = false

# Minimum execution time before logging
console_log_threshold = 0

# Decimal precision for displayed times
precision = 2

# Time threshold for slow node highlighting
slow_time = 10

# Frontend colors
active_color = "#4fc3f7"
badge_color = "#1e7832"
badge_color_slow = "#962828"