ComfyUI Extension: comfyui-metadata-nodes

Authored by DarkStarSword

Created

Updated

0 stars

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

ComfyUI custom nodes for preserving and extracting workflow metadata

Looking for a different extension?

Custom Nodes (2)

README

ComfyUI Metadata Nodes

A ComfyUI custom node extension that preserves and extracts workflow metadata from PNG images.

Features

  • Load Image with Metadata: Loads images from disk while preserving ComfyUI workflow metadata embedded in PNG files.
  • Extract Metadata Text: Extracts text fields from supported text nodes including CLIP Text Encode (Prompt), Display Any (rgthree), and Show Any (comfyui-easy-use).

Installation

  1. Copy this entire folder to your ComfyUI custom_nodes directory.
  2. Restart ComfyUI or reload custom nodes.
  3. Use Load Image with Metadata in place of the built-in Load Image node when loading images with embedded metadata.

Usage

Loading Images with Metadata

Use Load Image with Metadata instead of Load Image:

  1. Load a PNG file that contains ComfyUI workflow metadata.
  2. The node outputs both the image tensor and the workflow metadata string.
  3. Connect the metadata output to an Extract Metadata Text node.

Extracting Text

Use Extract Metadata Text to extract text from supported text nodes:

  1. Connect the workflow_metadata output from Load Image with Metadata.
  2. Optionally specify a node_id (defaults to 0 for auto-detection).
  3. The node outputs the extracted text.

Supported text nodes:

  • CLIP Text Encode (Prompt)
  • Display Any (rgthree)
  • Show Any (comfyui-easy-use)

Node IDs

When multiple supported text nodes are present, you'll need to specify the node ID manually. The error message will show the discovered node IDs.

License

MIT

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Learn more