FireRed Image online editor interface and product branding

ComfyUI workflow guide

FireRed Image Edit ComfyUI: Official Workflow Setup

Install the official FireRed Image Edit 1.0 workflow, connect the correct model files, and resolve the errors that commonly appear on first launch.

Official 1.0 workflow · Local node graph · Version notes for 1.1

What this guide covers

Key point 01

Official workflow

The downloadable JSON comes from the FireRedTeam 1.0 ComfyUI repository on Hugging Face.

Key point 02

Correct model mapping

Keep the 1.0 workflow paired with compatible 1.0 assets instead of assuming a 1.1 checkpoint is a drop-in replacement.

Key point 03

Local control

Use ComfyUI when you need a reusable node graph, local files, and repeatable image-editing runs.

Before installation

Know which FireRed package you are loading

FireRedTeam publishes a dedicated ComfyUI package for version 1.0. The 1.1 model has its own model page, but the team does not label the 1.0 workflow JSON as an official 1.1 workflow.

Workflow file

firered-image-edit-1.0.json stores the node graph and parameter wiring, not the model weights.

Model assets

Download the files referenced by the official repository and place them in the matching ComfyUI model directories.

Hardware

Local requirements depend on checkpoint precision, offloading, image size, and the rest of your graph. Start with a modest resolution.

Installation

Set up the FireRed Image Edit workflow in five steps

Use the original source links and preserve the workflow-model version pairing while you test the first image.

  1. 1

    Update ComfyUI

    Use a current ComfyUI installation, start it once, and confirm that its manager and model folders are available.

  2. 2

    Download the official JSON

    Save firered-image-edit-1.0.json from the FireRedTeam ComfyUI repository; do not copy a preview page as HTML.

  3. 3

    Install the model files

    Follow the repository file list and put every checkpoint, encoder, and VAE in the directory expected by its loader node.

  4. 4

    Import and reconnect

    Drag the JSON into ComfyUI. If a loader is red, select the locally installed file instead of changing unrelated nodes.

  5. 5

    Run a small test

    Upload a clear source image, write one specific edit instruction, use a fixed seed, and queue a low-resolution first run.

Compatibility

Official 1.0 ComfyUI workflow versus FireRed 1.1

Both belong to the FireRed Image Edit family, but their published packaging is different. Choose by workflow support, not only by version number.

Compatibility1.0 ComfyUI package1.1 model release
Official workflow JSONPublished by FireRedTeamNo separate official 1.1 JSON is linked from the 1.0 package
Best starting pointVisual local node workflowLatest model research and custom integration
Version pairingDesigned around 1.0 assetsTreat as a separate model until compatibility is verified
Setup styleImport JSON and map loadersFollow the model card and repository instructions
Official source1.0-ComfyUI on Hugging Face1.1 model page on Hugging Face

Direct download

Fetch the workflow without saving the Hugging Face page

The resolve URL returns the raw workflow JSON. Open the downloaded file and confirm that it begins with JSON before importing it.

  • Keep the original filename so it is easy to compare with upstream updates.
  • The JSON is only the graph; model weights must be downloaded separately.
  • Check the source repository again before replacing loader nodes or model names.
Terminal
curl -L \
  -o firered-image-edit-1.0.json \
  https://huggingface.co/FireRedTeam/FireRed-Image-Edit-1.0-ComfyUI/resolve/main/firered-image-edit-1.0.json

Troubleshooting

Fix common ComfyUI import and runtime errors

Most first-run failures are caused by a saved web page, a missing custom node, a wrong model path, or insufficient memory.

The JSON will not import

Open the file in a text editor. If it starts with HTML, download the raw resolve URL again.

A node is red or missing

Read the node name, install only the required extension through a trusted source, restart ComfyUI, and reload the graph.

The loader cannot find a model

Place the file in the model type folder used by that loader, refresh model lists, and select the local filename.

Out-of-memory during generation

Reduce image dimensions and batch size, close other GPU tasks, and use supported offloading or lower-precision options.

Primary sources

Verify downloads against official FireRedTeam pages

These sources distinguish the published 1.0 ComfyUI workflow from the model releases.

FAQ

FireRed Image Edit ComfyUI questions

Is there an official FireRed Image Edit ComfyUI workflow?

Yes. FireRedTeam publishes a FireRed Image Edit 1.0 ComfyUI repository with the firered-image-edit-1.0.json workflow.

Does the official workflow support FireRed Image Edit 1.1?

The published JSON is explicitly named for 1.0. Do not assume it supports the 1.1 model without checking the 1.1 model card and node compatibility.

Does the workflow JSON include model weights?

No. It stores the node graph and settings. The model files are separate downloads and must be placed in ComfyUI model folders.

Why are nodes missing after import?

Your ComfyUI install may not include every node type used by the workflow. Identify the exact missing node and install its trusted extension before retrying.

Can I run FireRed Image Edit without ComfyUI?

Yes. The official GitHub repository documents a Python inference path, and the online editor provides a browser option without local setup.

Where should I report workflow issues?

First reproduce the problem with the unmodified official workflow, then use the issue or discussion channel linked by the FireRedTeam repository.

No local setup

Try FireRed Image Edit in your browser

Use the online editor when you want to test an instruction before building a local ComfyUI graph.