WATCH VIDEO TUTORIAL:
This includes the GGUF quantized Flux Fill Dev Model, the FLUX 1 Redux Dev Model, the SIG CLIP Vision Model, and the Flux Turbo Alpha Lora Model for faster generation. We’ll also guide you through placing these files in the correct folders so that you can jump right into image inpainting without any hiccups. By ensuring the models are properly set up, you avoid any compatibility issues and can immediately begin exploring all the amazing features that Flux Tools Fill Dev and Flux Redux have to offer.
In the Flux Tools Fill Dev portion of the workflow, you’ll discover how to effortlessly edit and inpaint images by applying detailed, mask-based modifications to specific areas of a photograph. For instance, if you want to change someone’s hair color, all you need to do is type in “Hair” as your prompt, upload your reference image, and watch as the model identifies and modifies only the hair region, leaving the rest of the image intact. This laser-focused editing is perfect for creators looking to refine small details without affecting the original image composition.
Next, we’ll dive into the Flux Redux Model, an incredible addition by Black Forest Labs that allows you to generate variations of your images. By combining Flux Redux with the Flux Tools Fill Dev Model, you can introduce controlled variations in other image aspects while maintaining the core features of your subject. For example, if you’ve successfully changed someone’s hair color and want to explore different outfit styles or background changes, Flux Redux can seamlessly create fresh, imaginative results without compromising the quality of your base image.
This workflow isn’t just limited to human subjects; you’ll see how well these models handle non-human images, such as houses, cars, or any other object. By simply entering the relevant prompt (e.g., “house” or “car”) and uploading a reference image, you’ll be able to morph, restyle, and recolor any aspect of the scene. Want to replace a plain background with a stylish runway? No problem—Flux Tools Fill Dev and Flux Redux work together perfectly to make it happen.
We’ll also cover helpful troubleshooting tips. If you encounter any issues with nodes not working, we’ve got you covered with a simple walkthrough on how to use the ComfyUI Manager to install missing custom nodes or update them. After a quick ComfyUI restart, you’ll be back on track and ready to continue exploring the endless possibilities of inpainting.
Load CLIP Vision: https://huggingface.co/Comfy-Org/sigclip_vision_384/blob/main/sigclip_vision_patch14_384.safetensors
Load Style Model: https://civarchive.com/models/969352/flux-adapter-flux1-redux-dev
Load Flux Fill Dev GGUF Model: https://huggingface.co/YarvixPA/FLUX.1-Fill-dev-gguf/tree/main
Load Lora Model: https://huggingface.co/alimama-creative/FLUX.1-Turbo-Alpha/tree/main
Description
Version 1.0
FAQ
Comments (12)
Absolutely great wf. Please add a switch for manual mask option in v2.
ok, but its easy to do it, just remove the florence2 and Sam2segmentation and conect the growmask to the load image directly and using the open in Mask Editor, you can make the image yourself and that will work.
here is the Ultimate Flux Fill Inpainting Redux Workflow with Manual Mask Applying: https://file.io/Y2yw5wKo3puy
@AIMotionStudio Ahh thank you. Will try that.
@AIMotionStudio Apologies, I was offline for the Holiday and missed this link. File has been deleted, it says.
@AIMotionStudio I added an image/mask switch from Impact Pack. Works perfect for switching between the FlorenceSAM-2 mask and manual. Thanks man. I was a little confused as to how to do that before your tip. Appreciate it!
I'm running into the following issue:
# ComfyUI Error Report
## Error Details
- Node ID: 69
- Node Type: DownloadAndLoadFlorence2Model
- Exception Type: OSError
- Exception Message: Error no file named pytorch_model.bin, model.safetensors, tf_model.h5, model.ckpt.index or flax_model.msgpack found in directory C:\Users\alelks\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\models\LLM\Florence-2-Flux-Large.
## Stack Trace
Just seems to change color. Doesn't actually change the shape of objects.
Workflow freezing comfyui mate. Any idea what's going on?
My ComfyUI is freezing on loading this workflow
My ComfyUI is freezing on loading this workflow
My ComfyUI also freezing when I try and load this workflow. (shows partial view of workflow... have downloaded ClipV, style, GGUF, Lora etc)










