Skip to main content
Browse 20 learning paths

AI learning path

ComfyUI Workflows

Build node-based image workflows that are reproducible, portable, and easier to debug than a growing pile of screenshots.

For Technical image creators

6 original lessons on ToolDix, roughly 100 minutes end to end

Start the course

Course outline

  1. 1Read a ComfyUI Workflow GraphUnderstand the roles of model, prompt, latent, sampler, decode, and save nodes before changing a workflow.Beginner17 min
  2. 2Debug a ComfyUI Workflow SystematicallyDiagnose missing nodes, model mismatch, memory failures, and unexpected outputs by reducing a graph to a known-good baseline.Intermediate18 min
  3. 3Build a Minimal Text-to-Image WorkflowWire the seven-node spine from an empty canvas, learn what each sampler setting does by measuring it, and keep a baseline you can extend safely.Beginner17 min
  4. 4Add Image Input and Inpainting to a WorkflowExtend the spine into image-to-image and masked editing, wire the mask path correctly, and understand why an inpaint that ignores your mask is usually a wiring problem.Intermediate17 min
  5. 5Build a Two-Stage Upscale WorkflowGenerate at native resolution and refine at delivery size, wire the second pass so it improves detail without reinventing the image, and keep tiling from spawning duplicate subjects.Intermediate16 min
  6. 6Package a Workflow Someone Else Can RunShip a graph with the dependency manifest, model list, and safety context it needs, and treat an incoming workflow as code from a stranger rather than as a document.Intermediate15 min

Optional, after the lessons

Further reading from primary sources

Published by other organizations and kept here for reference. The lessons above are the ToolDix course; these are where to go once you want the vendor's own documentation.

5 resources