Skip to content
Help Build a Better ComfyUI Knowledge Base Become a Patron
Tutorial SeriesComfyUI Advanced Tutorial2.3 FLUX.1 Fill

Kunlun Wanwei Releases SkyReels-V2 Infinite-Length Film Generative Model

04/21/2025

Flux Fill Workflow Step-by-Step Guide

Flux fill dev.1

Flux.1 Fill Dev is an open-source image editing model released by Black Forest Labs, a key component of the FLUX.1 Tools suite, focused on image inpainting and outpainting tasks.

In this tutorial, we will briefly introduce the model and provide complete instructions for Flux Fill model installation, workflow files, and usage.

In this guide, we will cover:

  • Flux.1 fill dev inpainting workflow
  • Flux.1 fill dev outpainting workflow

Introduction to Flux.1 Fill dev Model

Flux.1 Fill dev is one of the core tools in the FLUX.1 Tools suite released by Black Forest Labs, designed specifically for image repair and extension. This model is primarily used for:

  • Image inpainting: filling in missing or removed areas of an image
  • Image outpainting: seamlessly extending the boundaries of existing images
  • Precise control over generated content using masks and prompts

Key features of Flux.1 Fill dev:

  • Powerful image inpainting and outpainting capabilities
  • Excellent prompt understanding and following ability, accurately capturing user intent while maintaining high consistency with the original image
  • Advanced guided distillation training techniques, making the model more efficient while maintaining high-quality outputs
  • Flexible usage license, generated content can be used for personal, scientific, and commercial purposes - please check the FLUX.1 [dev] Non-Commercial License

Flux.1 Fill dev model repository: Flux.1 Fill dev

Flux.1 Fill dev Complete Workflow

πŸ’‘

Before starting this section, please ensure you have upgraded to the latest version of ComfyUI. You can refer to How to Update and Upgrade ComfyUI for the update process.

The workflow images in this article include model file downloads and workflow information:

  • Simply drag the corresponding image into ComfyUI to load the workflow
  • If automatic downloads fail, please refer to the manual installation section to complete model installation

Since the Flux.1 Fill dev model requires agreement to the corresponding model repository license, please ensure you have agreed to the license before starting, so you can download the corresponding workflow.

Repository License

Flux.1 Fill dev Inpainting Workflow

1. Inpainting Workflow File Download

Download the image below and drag it into ComfyUI to load the workflow

ComfyUI Flux Fill Inpainting Workflow

We will use the following image as our input:

ComfyUI Flux Fill Inpainting Input Image

πŸ’‘

This image has already been processed and contains an alpha channel, so you don’t need to process it further. If you want to process it yourself, you can use the original image below as an input example:

ComfyUI Flux Fill Inpainting Original Image

And refer to the ComfyUI Mask Editor Usage Tutorial section to complete the mask drawing.

2. Manual Model Installation

If you encounter problems with model downloads, please refer to the model file list below, manually download the corresponding model files, and place them in the appropriate ComfyUI directories.

You need to download the following model files:

Model NameFile NameInstallation LocationDownload Link
CLIP Modelclip_l.safetensorsComfyUI/models/text_encodersDownload
t5xxl_fp16.safetensorsComfyUI/models/text_encodersDownload
VAE Modelae.safetensorsComfyUI/models/vaeDownload
Flux Fill Modelflux1-fill-dev.safetensorsComfyUI/models/diffusion_modelsDownload

File storage locations:

ComfyUI/
β”œβ”€β”€ models/
β”‚   β”œβ”€β”€ text_encoders/
β”‚   β”‚    β”œβ”€β”€ clip_l.safetensors
β”‚   β”‚    └── t5xxl_fp16.safetensors
β”‚   β”œβ”€β”€ vae/
β”‚   β”‚    └── ae.safetensors
β”‚   └── diffusion_models/
β”‚        └── flux1-fill-dev.safetensors

3. Check the Corresponding Nodes and Complete the Workflow Run

ComfyUI Flux Fill Inpainting Workflow Step Diagram

  1. Ensure flux1-fill-dev.safetensors is loaded in the Load Diffusion Model node
  2. Ensure the following models are loaded in the DualCLIPLoader node:
    • clip_name1: t5xxl_fp16.safetensors
    • clip_name2: clip_l.safetensors
  3. Ensure ae.safetensors is loaded in the Load VAE node
  4. Upload the input image provided in the document to the Load Image node; if you’re using the version without a mask, remember to use the mask editor to draw the mask
  5. Enter the content you want for the masked part of the image in the CLIP Text Encode(Positive Prompt) node
  6. Click the Queue button, or use the shortcut Ctrl(cmd) + Enter to run the workflow

4. Workflow Explanation

This version is a complete inpainting workflow. Due to the Flux series models’ excellent prompt understanding capability, we only need to input simple prompts to get very good results.

If you’re a beginner using an inpainting workflow for the first time:

  • We uploaded the input image in the Load Image node and used the MaskEditor tool to draw the mask, which marks the area that needs to be modified by the model

Flux.1 fill dev Outpainting Workflow

1. Outpainting Workflow File Download

Download the image below and drag it into ComfyUI to load the workflow

ComfyUI Flux Fill Outpainting Workflow

We will use the following image as our input:

ComfyUI Flux Fill Outpainting Input Image

2. Manual Model Installation

The model installation is the same as the inpainting section, please refer to the inpainting section above.

3. Check the Corresponding Nodes and Complete the Workflow Run

ComfyUI Flux Fill Outpainting Workflow Step Diagram

  1. Ensure flux1-fill-dev.safetensors is loaded in the Load Diffusion Model node
  2. Ensure the following models are loaded in the DualCLIPLoader node:
    • clip_name1: t5xxl_fp16.safetensors
    • clip_name2: clip_l.safetensors
  3. Ensure ae.safetensors is loaded in the Load VAE node
  4. Upload the input image provided in the document to the Load Image node
  5. In the Pad Image for Outpainting node, you can customize the size of the areas you want to extend in each direction
  6. Enter the corresponding description in the CLIP Text Encode(Positive Prompt) node
  7. Click the Queue button, or use the shortcut Ctrl(cmd) + Enter to run the workflow