I/OLimited Capture

Load Image

class_type: LoadImage

Source repo

Loads an image from the ComfyUI input directory for use in img2img, ControlNet, or other image-based workflows. Records only the filename, creating a fragile provenance link to the source material.

What It Does

LoadImage reads an image file from the ComfyUI input directory and converts it into an IMAGE tensor for downstream processing. It also optionally outputs a MASK for images with alpha channels.

This node represents a significant provenance gap: it records only the filename of the loaded image, not its content hash or origin. If the file is moved, renamed, or modified, the provenance link is silently broken. For img2img workflows, the source image is arguably as important as the prompt—yet it is the least well-preserved piece of metadata.

Inputs

imageSTRING

Image filename from the input directory.

Outputs

IMAGEIMAGE

Loaded image tensor.

MASKMASK

Alpha channel mask (if present).

What Numonic Captures

  • Input image filename

Known Gaps

  • Image content hash — file could change without detection
  • Image origin — was it AI-generated, photographed, or downloaded?
  • Image metadata — EXIF/XMP from the source file is not propagated

Related Nodes

Save ImageI/O
VAE EncodeImage Processing
Apply ControlNetControl
Back to ComfyUI Nodes

Capture ComfyUI metadata automatically

Numonic extracts workflow metadata from every ComfyUI generation — models, samplers, seeds, prompts, and custom nodes. Track provenance, maintain compliance, and never lose a workflow.

Return to the Guide