Hello!
Version 2511 v1:
This workflow works fine with 8GB VRAM, if you have a better card, you might want to disable the Lightning LoRA, add more steps, and cfg to every KSampler. Maybe 20 steps and cfg 2.5? You should try. Unlike the previous versions of this workflow, this has a custom node and uses a Q version.
Resources:
[Model] Qwen-image-edit-2511-Q2_K.gguf:
[CLIP] qwen_2.5_vl_7b_fp8_scaled
[VAE] qwen_image_vae
[Nodes] Comfyui-QwenEditUtils
About the prompts:
These prompts are just examples, so you might want to change them, because, depending on the original image they may work or may not work. Also, I totally recommend translating the prompts to Simplified Chinese, because in that language they will have more adherence. For example: First prompt is orbit the camera 45 degrees. Second prompt is right-side. Third prompt is zoom out. You can use English too, but, you'll need better luck.
_____________________________
Previous version text:
Hello!
People asked me to update this workflow, so, here it is! This version uses:
And 8 steps LoRA:
Also, I added a secondary image, if you want to add something to your new images, if you don't want, you can disable the load node. Remember always to change the prompts to do what you need.
The file you'll download has the JSON file as well as a PNG, so you can use either in ComfyUI
Cya!
_____________________________
Previous version (1.0):
Hello!
I'm kinda new to ComfyUI, so I made this spaghetti nightmare, this will create 6 different images from one image using Qwen Image Edit with 4 steps each one (I'm low VRAM!).
Files used:
Model: qwen_image_edit_fp8_e4m3fn
Lora: Qwen-Image-Lightning-4steps-V1.0
Clip: qwen_2.5_vl_7b_fp8_scaled
Vae: qwen_image_vae
Have fun!
Description
Now uses Qwen Edit 2511 (Q2), its proper LoRA, and QwenEditUtils
FAQ
Comments (14)
it makes three images that are identical to the source image, changing the prompt doesnt do anything
Sometimes it does that, for unknown reasons, but not always. My guess is something to do with the Lightning LoRA and low steps. Do this, disable 2 of the 3 outputs, and the lightning LoRA, and in KSampler of the enabled output add to 20 steps and cfg to 2.5, and try it, if it works, you can do the same with the other 2 output's KSamplers, but it will take a while.
I'm getting some sort of ghosting within my images. Sometimes it's one out of the three. Other times it's all of them. Basically, it's like I can see a faint "bleed through" of the original reference image. Any thoughts?
Ghosting?... No idea.
@gabrielx The best way I can describe it is it's like a double exposure of the original reference image and created image.
@MysticMindAi you have the wrong models packed up. make sure you match the exact ones, if you use e.g. qwen image instead of image edit 2511 it will show that.
@iamunidentifiedn1614 I've double checked the models, and everything is good. In fact, it wouldn't work at all if the models were wrong. So, then i tried a workflow with the new multiangle nodes, which is essentially what this workflow is without them, right? While it works im trying to compare the two workflows besides those nodes and the usage of the lora that's needed to make it work. There are a few other differences thinking
I'll report my findings later.
у меня было то же самое. обновите ComfyUI
how to make it uncensored
not working, trash
Hey there I'm new but when I upload the workflow the " GGUF LOADER" doesn't let me run. where do I need to conect it? thanks
Oh FKN Outstanding! - Using NSFW AIO v23 - running on BlockFlow - with RTX Pro 6000 on RunPod severeness. > I2V- LRX 1.3 >Muffins Flat2VR = The birth of a new Incel! ... Bye Bye Reality!