Fastest Flux Canny Model Workflow EXPOSED by AI Expert!
Added 2025-01-12 14:30:25 +0000 UTCBest Flux Canny workflow ever! Deep tutorial with incredible details about concept of canny itself and workflow insane capabilities!
Download workflow Basic (V1) : https://drive.google.com/file/d/12Vd6E81-pdFfv04Oci1dMJsGrUtJfdEX/view?usp=sharing
Download Advanced Workflow (V2) : Attached Bellow
🇷🇺 From Russia you can use this link For V2 workflow :
Advanced Workflow (V2) Details :
1- It has complete control for set the image quality (very low to 8K) even it eliminates the need for upscaling!
2- It gives you the ability to control the influence of Canny, for example, 20%, 40%, or more or less which increases Flux flexibility to generate more creative images.
3- Fast Generation speed without compromising the quality of the final output
Feel free to download it from below, try it out, and send me the results you get.
Models Links :
Flux-Canny model and VAE : https://huggingface.co/black-forest-labs/FLUX.1-Canny-dev/tree/main
Put it here : \ComfyUi New\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\models\unet
.
Flux 8 step Turbo Lora : https://huggingface.co/alimama-creative/FLUX.1-Turbo-Alpha/tree/main
Put it here : \ComfyUi New\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\models\loras
.
Dual Clip : https://huggingface.co/zer0int/CLIP-GmP-ViT-L-14/tree/main
Put them here : \ComfyUi New\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\models\clip
.
Flux1 Dev-fp8 and Vae : https://huggingface.co/XLabs-AI/flux-dev-fp8/tree/main
Put it here : \ComfyUi New\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\models\unet
.
- Complete Guide for Beginners (watch videos below on by one) :
1-Install ComfyUi and Flux Locally : https://www.youtube.com/watch?v=txDFK-RcUq4&t=28s
2-Guide for Low-end systems for Flux : https://www.youtube.com/watch?v=GtPdEwVwtnM&t=906s
3-How to Create ai images with your own face : https://youtu.be/JeYmXYRYZ8k
.
Follow me on social media :
Patreon : https://www.patreon.com/c/Jockerai
Youtube : https://www.youtube.com/@Jockerai
Civitai : https://civitai.com/user/jockerai
Telegram : t.me/jokerrai
.
Comments
I’m experiencing the exact same issue. The first image generates fine, but the second one is just black. I’ve double-checked all the parameters, watched all the previous installation videos, and even re-watched the workflow tutorial three times, but I still can’t figure out what’s causing the problem. The same thing happens when I try using the depth map workflow as well.
Geziel Machado
2025-01-26 20:40:52 +0000 UTCSadly, it just doesn't work for me. The workflow produces a blank (black) image every time. It successfully creates the initial image with canny but fails to create a final render... it is always just pure black with nothing.
Adam Frangione
2025-01-17 23:18:39 +0000 UTCthank you very much :D
Saul Mendoza Martinez
2025-01-15 17:01:01 +0000 UTCHey bro. I used both this is the download link of the lora : https://civitai.com/models/199968/liquid-water-flow-style-fluxsdxl15?modelVersionId=890720 . Use 1.2-2 for strength of the lora
moh legato
2025-01-15 17:00:21 +0000 UTCHi bro, in the final result of your ice style miniature. Did you use a lora or just promt? or both?
Saul Mendoza Martinez
2025-01-15 16:53:04 +0000 UTCyes try using %0 for canny percentage. Each lora impact to it's group and not to other so it's up to you use lora in both or just one or none.
moh legato
2025-01-15 14:34:44 +0000 UTCIs it possible to remove the image creation in the first group completly? Also has the lora in the first group any impact on the second group?
Danny
2025-01-15 13:01:36 +0000 UTCThank you. Where are the questions 🤔
moh legato
2025-01-15 12:57:33 +0000 UTCOk cool it is a great workflow thats for sure. Could you only answer the last 2 questions, then I am good :)
Danny
2025-01-15 12:54:20 +0000 UTCFor end step it should be at least equal to amount of main step value and not less (25 step that you want). Doesn't matter if it set to higher value. Anyway it will stop the process at 25. I recommend set to 999 and leave it alone and don't touch it again😁
moh legato
2025-01-15 12:49:33 +0000 UTCAlso has the lora in the first group any impact on the second group?
Danny
2025-01-15 12:33:27 +0000 UTCok so the end steps have to match the steps in the first group? Because in your version is it 8 to 40 My goal is to remove the turbo lora,so I set the steps to 25 in the first group - second group also 25 or leave at 40? Is it possible to remove the image creation in the first group completly?
Danny
2025-01-15 12:33:04 +0000 UTCYes sure. You must have same amount of steps in both. For that reason we just have one node to set the steps. Indeed it's a single process that some parts executed by canny and rest executed with flux. That "End at step" options means when the flux should stop! We say for example 40 but since we have 8 steps for whole workflow flux will stop processing at 8. If you increase steps for example to 50 steps you need to set end at step to 50.
moh legato
2025-01-15 12:29:26 +0000 UTCThe steps in canny settings are for the entire workflow? To create the image in canny settings and in the group "flux"? Because the sampler in the second group dont have steps only end at step, could you explain this a bit further? Thank you :)
Danny
2025-01-15 12:19:44 +0000 UTCah its there now...i must not of been logged in correctly or something.... apologies !!
Thomas Englefield
2025-01-12 19:52:21 +0000 UTCNo I don't think so 😁 in Patreon posts the download link is always at the very bottom, where description ends. :)
moh legato
2025-01-12 19:50:14 +0000 UTCHello there . What do you mean? Is it a kind of special image or something?🤔
moh legato
2025-01-12 19:48:41 +0000 UTCAm I going blind? I cannot for the life of me find the link for V2 download :-)
Thomas Englefield
2025-01-12 19:47:06 +0000 UTCHi friend. Can you make a video about How to Create Viral America's Got Talent which is very trendy these days?
kadir alpago
2025-01-12 19:39:55 +0000 UTC
