EasyCascadeKsampler(samplerCascadeSimple)

Description

The EasyCascadeKsampler node is a simple sampler wrapper designed for Cascade-style pipelines in ComfyUI-Easy-Use. It accepts a PIPE_LINE — the Easy-Use pipeline object that bundles model, conditioning, and latent state — performs sampling, and returns both the updated pipeline and the generated image. It also integrates with Easy-Use's image preview, save, and Sender utilities, making it a convenient single-node sampling step for Cascade workflows.

Inputs

• pipe (PIPE_LINE, required): The Easy-Use pipeline containing the model, positive/negative conditioning, latent, and other sampling data.
• image_output (COMBO, required): Controls how the generated image is handled. Default: Preview. Available options:
  • Hide — Do not preview or save the image.
  • Preview — Display the image in the UI.
  • Save — Save the image with the configured prefix.
  • Hide&Save — Save without showing a preview.
  • Sender — Send the image to a configured destination using the link ID.
  • Sender&Save — Send the image and also save it.
• link_id (INT, required): The destination link ID used when image_output is set to Sender or Sender&Save.
• save_prefix (STRING, required): The filename prefix used when the output is saved. Default: ComfyUI.
• model_c (MODEL, optional): An optional model input. This can be used to supply a separate Cascade model or override the model contained in the pipeline when needed.

Outputs

• PIPE_LINE: The updated pipeline after sampling, ready to be passed to downstream nodes.
• IMAGE: The generated image tensor, available for preview, saving, or further image processing.

Usage Notes

This node is intended for Cascade-style sampling workflows where the pipeline object carries all necessary components. In most cases you can simply connect a PIPE_LINE source, choose an image_output behavior, and use the output image or pipeline as needed.

Use link_id in combination with the Sender modes to route images to another part of your workflow or external tools. For saved outputs, adjust save_prefix to control filenames. The optional model_c input is useful when you need to pass a model independently of the pipeline, such as when mixing Cascade stages or using a separate control model.