ComfyUI-Wanify: Adaptive Image Resize Node

This custom node for ComfyUI provides adaptive image resizing based on target pixel counts, maintaining aspect ratio and supporting different quality levels. It is useful for workflows that require images to fit within specific pixel budgets (e.g., for video, AI models, or memory constraints).

Features

• Resize images to target pixel counts (720p/480p, high/medium/low quality)
• Maintains aspect ratio
• Outputs resized image, width, height, and total pixel count
• Fast and quality modes

Installation

1. Download or clone this repository:

   git clone https://github.com/blird/ComfyUI-Wanify.git
   

2. Copy the comfyui-wanify folder into your ComfyUI/custom_nodes/ directory.
3. Restart ComfyUI. The node will appear as "Adaptive Image Resize" under the image/transform category.

Usage

• Inputs:
  • image: Input image tensor (ComfyUI format)
  • model: Target resolution model (720p or 480p)
  • quality: Quality setting (high, medium, low)
  • fast_mode: Boolean, use fast interpolation (default: True)
• Outputs:
  • image: Resized image
  • width: Output width
  • height: Output height
  • total_pixels: Output width * height

Example

Adaptive Image Resize
  ├─ image: (input image)
  ├─ model: 720p
  ├─ quality: high
  └─ fast_mode: True
Outputs:
  ├─ image: (resized image)
  ├─ width: 1280
  ├─ height: 720
  └─ total_pixels: 921600

Requirements

• ComfyUI
• PyTorch (ComfyUI default)