Aitrepreneur
patreon
QWEN COMBO ULTRA WORKFLOW!
🕑 Added 2025-08-11 20:04:17 +0000 UTCEDIT 12 AUGUST: 4 STEPS LORA IS OUT! THE WORKFLOW AND INSTALLERS HAVE BEEN UPDATED! SO NOW YOU CAN GENERATE AN IMAGE TWICE AS FAST...AI GOES REALLY FAST HUH..😅
YOU CAN REDOWNLOAD THE WORKFLOW AND EITHER USE THE QWEN-ULTRA-MODELS-NODES_INSTALL.bat inside the comfyui folder or dl the lora from this link: https://huggingface.co/Aitrepreneur/FLX/resolve/e70ebc249a07b2076db848b166d8ff1c759ce953/Qwen-Image-Lightning-4steps-V1.0.safetensors?download=true
and put it inside the ComfyUI_windows_portable\ComfyUI\models\loras folder
HAVE FUN! 🙃
Hey everyone! I've just released a new video on the NEW QWEN IMAGE MODEL!
THE NEW BEST TEXT TO IMAGE MODEL ever made!
Insane artstyle, incredible prompt following, great text design and just amazing quality. It can do everythingl. BUT...NOT BY ITSELF...
The only way to TRULY Make it the best is THE SUPER COMBO TEAM QWEN + WAN 2.2! And My Workflow makes this possible!
You can check out the video right here: https://youtu.be/7P4LHEAEGNg
I also created a 1-click installer for the Webui for my Patreon supporters for Local AND Runpod, that you can find right here: https://www.patreon.com/posts/136261655
Here the link to the manual install guide: https://pastebin.com/18cv3Ww7
Down below you will also find the QWEN_COMBO_ULTRA_WORKFLOW.json for free for you to use!
As always, supporting me on Patreon allows me to keep creating helpful resources like this for the community. Thank you for your support - now go have some fun😉!
Comments
Airdropkazan
mat1 and mat2 shapes cannot be multiplied (784x1280 and 3840x1280) i tried several clip models but it didnt work for me. google says maybe its from the node itself.
Shelby Black
How do you fix the white screen?
MuseEngine
i managed to make all the issues and alert work by using chatgpt. DO the same if you have issues
Josh inmn
I recall someone asking for an auto prompter but I cant recall where someone in here though. I created an wan2.2 prompt generator to properly generate the correct style of prompts for the most realistic, and advance. It uses either LMStudio or OLLAMA for the back end. I have an exe and the source code on my github. Hope it helps someone. https://github.com/Wontfallo/Wan2-2_Prompt_Generator.git
Paul Warren
I'm getting this error, can anyone advise? mat1 and mat2 shapes cannot be multiplied (5292x1280 and 3840x1280) thanks guys
Zoran Kanti-Paul
I tried the Qwen image to image, every time I do it either comfyui just freezes up or my pc actually crashes.
Josh Gmit
different models need to be handled differently. You may get a workflow for SDXL or other Model-Architectures on CivitAi too.
BigHomie
Hi, where in this workflow would i add the Juggernaut XL from CivitAi ? or generally, how do i modify this workflow for other LoRAs and Models from CivitAi ????????
Bob
i think i found best seed: 999999999999999 1039338885633958
Aitrepreneur
yes you can create 2 int nodes and then connect them to the width and heigh part of the nodes
Mikael Svenson
Is there a way to share height/width between the two parts of the combo? To save time updating both t2v and i2v when changing aspects?
Gardan
Check his previous youtube video for WAN 2.2, there you have the workflow for image to video, and you can copy that part over in comfyui (ctrl+drag to select that part and ctrl+c, ctrl+v in the new workflow)
HealingPaint
Would it be possible to add one more module to this workflow to complete it; turn your final image into a video, like runi it through wan 2.2? That would be awesome.
Gardan
Thank you so much, I got it working after getting only blurry images (like I got also with the WAN 2.2 Image2Video). I finally figured out that I was missing some Loras and now it works. Thanks so much for your work, you're awesome. Also actually this was probably your funniest video, with the comments about WAN 2.2. ;)
