How To Edit Any Image With FlUX Dev and FlUX Shnell in ComfyUI -Inpaint/Outpaint & Background Remove

Discover how to master inpainting/outpainting for low VRAM devices with Flux in comfyui.

Image manipulation Workflow:

Low Vram GGUF_NF4_FP8-16 Workflow:

Flux Upscale:

Install ComfyUI:

Hyper flux and Workflow introduction:

Flux Controlnet:

Model list.

NF4 Model:

Flux Dev GGUF:

Flux Dev fp8:

T5 GGUF Model:

T5 FP8 Model:
NB: you'll need to login or sign and agree to the terms of stabilityai to download the fp8 model

Clip VIT:

Flux hyper model Lora:

Flux dev add detail lora:

jasperai/Flux.1-dev-Controlnet-Upscalerr:

jasperai/Flux.1-dev-Controlnet-Depth:

jasperai/Flux.1-dev-Controlnet-Surface-Normals:

Controlnet Depth:

Contorlnet Canny:

Mistoline Controlnet:

Mistoline GitHub:

NB: The NF4 checkpoint loader, Mistoline Note and the perfection style Nodes has to be manually installed, Here are the steps:

Installing NF4 checkpoint loader:
open the manager click on install via git URL paste the following url and click on ok and wait till its done installing.

Installing perfection styler:
open the manager click on install via git URL paste the following url and click on ok and wait till its done installing.

Installing Mistoline:
open the manager click on install via git URL paste the following url and click on ok and wait till its done installing.

#comfyui #flux

[TheSneakyRobot]

ComfyUI Workflows And Models: 😁


Image manipulation Workflow:


Low Vram GGUF_NF4_FP8-16 Workflow:


Flux Upscale:

Install ComfyUI:

Hyper flux and Workflow introduction:

Flux Controlnet:

Model list.

NF4 Model:

Flux Dev GGUF:

Flux Dev fp8:

T5 GGUF Model:

T5 FP8 Model:
NB: you'll need to login or sign and agree to the terms of stabilityai to download the fp8 model

Clip VIT:

Flux hyper model Lora:

Flux dev add detail lora:







Controlnet Depth:

Contorlnet Canny:

Mistoline Controlnet:

Mistoline GitHub:

NB: The NF4 checkpoint loader, Mistoline Note and the perfection style Nodes has to be manually installed, Here are the steps:

Installing NF4 checkpoint loader:
open the manager click on install via git URL paste the following url and click on ok and wait till its done installing.


Installing perfection styler:
open the manager click on install via git URL paste the following url and click on ok and wait till its done installing.


Installing Mistoline:
open the manager click on install via git URL paste the following url and click on ok and wait till its done installing.

[bahethelmy]

Very nice video! Lots of work put there..

[generalawareness]

I can't get inpainting text onto an image.

[naserazmoon]

Hello, I would like to know if it is possible to replace a masked area of an image with my own image. For example, could I replace a painted-over area of clothing with my own clothing?

[andrzejsomiany]

Hi thank you for your flow is great.

I got an error that reads:

ImpactSwitch

Node 1404 says it needs input input1, but there is no input to that node at all

How can there be a cause?

[Wambalfa]

After installing the NF4 checkpoint loader via git the node is still red. Even a complete restart doesn't help🥲

[aelendys]

Awesome ! +1 subscriber !

As a total newbie in comfyUI, I must say it's a bit complicated but big thank you for your very clear explanations ! Juste to be sure, we can enable or disable each group nodes (Object Remover, SAM2 Inpainting (AUTO), etc.) but if I want to use SAM2 Inpainting (AUTO), the group "Inpainting (MANUAL)" is necessary (because it's here that the image is put), right !?

[mayankgupta]

its not working to execute the autopaint

i get the below error
ImpactSwitch
Node 1404 says it needs input input1, but there is no input to that node at all

for sam2 its able to recognize and mask, but there is no output image

[rifz]

Subd for including the workflow file! : ) thanks.
for future videos please don't add music. tutorials should not have music, it's so distracting from what is being said.

[startplayertwo]

Hey, can you make a workflow for flux style aligned reference sampler?

[AInfectados]

Why not include the CONTROLNET INPAINT model?

[profitsmimetiques]

the text to image prompt box does not work for me, in the complet eworflow

Failed to validate prompt for output 137:
* SDXLPromptStylerPreview 155:
- Value not in list: style: '{'content': 'base', 'preview': not in (list of length 107)
Output will be ignored
Failed to validate prompt for output 138:
Output will be ignored
Using pytorch attention in VAE
Using pytorch attention in VAE
Prompt executed in 0.47 seconds


everything is downloaded, and up to date



UPDATE : ok so in the prompt styler you must select "base" or it won't start.

Now i have this message : Node 1404 says it needs input input1, but there is no input to that node at all ... But it's selected as input 1 which is the text to image.

Thanks for your work but honnestly frustrating, always something not working and wasting hour just to try to make it work. Like here i've disabled everything except text 2 image, but this error keeps happening

Also, you should remove or change the prompt generator, it's not usable since it takes forever to load. Searge llm is better, but this and the image to prompt should be removed i guess. Everybody is using chatgpt for both prompt and image recognition, it's faster and more customizable, so keeping it just make the workflow complexe and not really useful