Update Highlights – Version 1.6
Completely redesigned to be more user-friendly, compact, and efficient, making the workflow easier to navigate while preserving its full flexibility.
The Dual Sampler and Hi-Res Fix pipeline have been reworked from the ground up, resulting in better image quality, improved coherence, and a more reliable generation process.
The entire Detailer section has been rebuilt with optimized models and a cleaner layout, providing more consistent face and hand refinement while remaining easier to configure.
The ControlNet section has also been fully revised, improving compatibility, organization, and overall usability for reference-based generations
Update Highlights – Version 1.5
Completely redesigned for better readability and ease of use, heavily inspired by vslinx’s workflow (shoutout to him).
The preview section has been revised for clearer and more intuitive feedback during generation.
The old dual-sampler setup has been replaced with a dedicated Hi-Res Fix node, improving image coherence and reducing anomalies during upscaling.
A new image filter now generates four base images, allowing users to select the most suitable one before moving forward with detailing.
Update Highlights – Version 1.4
Introduced a dedicated second sampler integrated with a Highres Fix script, significantly enhancing image fidelity and detail sharpness in final outputs. Expect smoother upscale transitions and richer textures, especially for close-ups or larger compositions.
Seamless integration of the new highres logic into the existing pipeline — no extra setup needed. Just plug and play for better results.
Further minor improvements to node connections and internal logic to ensure a more stable and optimized generation flow.
Update Highlights – Version 1.3
Added support for wildcards, allowing dynamic and flexible prompt generation
Introduced the ability to randomize wildcards for varied and automatic prompt creation across generations
Improved overall user experience with better layout clarity and workflow accessibility
Workflow remains modular and optimized for structured Multi-LoRA character and concept building
Update Highlights – Version 1.2
Refactored the "Adetailer" tabs for improved readability and usability
Added a control panel to easily toggle ON/OFF for various node groups
Globally improved layout structure and visual clarity for smoother navigation
Replaced and optimized samplers used within the "Adetailer" sections
Update Highlights – Version 1.1
Implemented a dedicated section for nipple detailing, with a toggleable node allowing users to enable or bypass the detail processing of each detailler section as needed.
Key Features:
▫ Structured Multi-LoRA Framework
LoRAs intelligently organized into dedicated sections (style, clothing, accessories)
Clear visual hierarchy for easy navigation and adjustments
▫ Precision Enhancement Tools
Optional Face Detailer module (separate controls for face, eyes, and hands)
Built-in Control Net integration for image-based generation
Full support for embeddings and A1111 prompt notation
▫ Optimized PonyXL Foundation
Pre-configured for maximum stability
Modular design allows customizing every aspect (poses, backgrounds, details)
▫ Professional-Grade Pipeline
Supports rapid iteration from concept to final render
Maintains consistency across generations
▫ Perfect For:
Professional character designers
Concept artists needing fast iteration
Style explorers wanting maximum control
▫ Download Includes:
Main workflow JSON
Pre-configured node templates
Ressources :
Detailers Models :
FACE
EYES
LIPS
NOSE
HANDS
NIPPLES
VAGINA
PENIS
Nodes used:
Join the Discord!
Come hang out, share your AI art, give feedback, ask questions, or just vibe with the community.
It’s also the best place for troubleshooting, commissions, and sneak peeks of upcoming work.
Description
FAQ
Comments (53)
Under the ControlNet node collection, I see on the Load Image node the following: NSFW_suspended001_bone_structure.png
What is this? Do I need it? And if so, where do I get it?
You just need to input any image that you want to use as a reference, like an OpenPose Image or juste a basic image !
This entire workflow works perfectly until I add a lora, adding a lora always results in the following error: Facedetailer 'NoneType' object has no attribute 'data'
That's really wierd can you be a little more specific? Where did you put the lora ?
@Mittoshura Character, style, any of those before the detailers, just fails every time.
@HaokiChan Hmm that's wierd because it doesn't look like it's an error from the Loras, did you try to desactivate all the detailers?
@Mittoshura I'm gonna have to do a bit of troubleshooting, because I've tried some other workflows besides this one after I was receiving the error, but I'm always somewhat disappointed with their hires fix nodes, they're just not as well-configured as yours is.
@HaokiChan That error just means FaceDetailer isn’t getting any image to work with.
Usually it’s one of two things:
LoRA mismatch → like trying to slap an SDXL LoRA on a 1.5 model. If the model fails to load, no image gets made → FaceDetailer ends up with None.
Broken wiring/order → adding a LoRA sometimes messes with the graph. If the sampler’s image output isn’t connected right, FaceDetailer basically runs on air.
Fixes:
Double-check your LoRA matches the base model (SDXL ↔ SDXL, 1.5 ↔ 1.5).
Try running the workflow with the LoRA but without FaceDetailer. If it still breaks, the LoRA is the problem. If it works, it’s just the wiring.
Update your FaceDetailer + LoRA Manager extensions (old versions love to throw this bug).
Re-connect the sampler’s image output to FaceDetailer manually, sometimes the link gets lost when you add nodes.
Keep me up with your troubleshooting so I can help you my best !
@Mittoshura Things have sadly changed in the past 2 weeks, with updates on ComfyUI and nodes, now I'm instead receiving this error when I generate just about anything, even without lora's:
'DifferentialDiffusion' object has no attribute 'apply'
@HaokiChan Are you using the latest version of the WF ?
@Mittoshura I just updated everything again, ComfyUI and all the node packages and everything seems to be working again, it's not throwing any errors, so that's good.
@HaokiChan Oh that's great news ! Hope you can finally have some fun with this WF hahaha
Tried to use this workflow on a fresh comfy install but cant get it to work at all, i cant get eden.art and ppm nodes to install
If you’re on a fresh ComfyUI install and can’t get eden.art and PPM nodes to work, here are the steps you can try:
Check the repos – those nodes are not bundled with Comfy by default.
Eden nodes → https://github.com/edenart-ai/ComfyUI-eden
PPM nodes → usually https://github.com/WASasquatch/PPM-ComfyUI (sometimes inside WAS suite).
Install them manually:
Go to your ComfyUI/custom_nodes folder.
Clone the repos:
cd ComfyUI/custom_nodes git clone https://github.com/edenart-ai/ComfyUI-eden git clone https://github.com/WASasquatch/PPM-ComfyUI
Or download the repos as ZIP and extract them into custom_nodes/.
Install dependencies:
Each repo may require extra packages. Check their requirements.txt.
From Comfy root folder:
pip install -r custom_nodes/ComfyUI-eden/requirements.txt pip install -r custom_nodes/PPM-ComfyUI/requirements.txt
Restart ComfyUI.
Sometimes you also need to clear cache: delete the ComfyUI\venv\Lib\site-packages\__pycache__ folders if errors persist.
@Mittoshura the eden nodes link doesnt work. just says page not found
Hello, trying to get this WF to work.
I get "Some nodes are missing":
CR Image Input Switch in subgraph 'Color Fix'
Is there some Node pack which needs a sprecific version or something like that?
Did you installe correctly all the node i listed ?
I installed missing nodes via ComfyUI Manager.
ComfyUI Impact Pack nightly 8.23.0
ComfyUI-Custom-Scripts nightly 1.2.5
efficiency-nodes-comfyui 1.0.8
Save Image with Generation Metadata nightly
Allor Plugin nightly
ComfyUI Impact Subpack 1.3.5
ComfyUI-ppm 1.0.43
ComfyUI_Mira nightly
Various ComfyUI Nodes by Type nightly
ComfyUI Image Saver nightly 1.15.2
Eden.art nodesuite nightly
cg-image-filter 1.6.1
ComfyLiterals nightly
rgthree-comfy 1.0.0
ComfyUI 0.3.60
ComfyUI_frontend v1.26.13
ComfyOrg EasyUse v1.3.1
ComfyUI-Manager V3.32.2
@Erunai Try to install the node that is bugging manualy via a git clone in your custome node folder
@Mittoshura Had to install Comfyroll Studio.
Now I'm missing the VAE in your pre Workflow.
Where can I get this?
lunaXLILNAIVAE_luna.safetensor
I only found a "lunaXLVAE_luna.safetensors" on hugginface.
I got it to work now. Had to scrap the Scheduler Selector and set scheduler manual for the KSampler's and fix a hidden node in the FaceDetailer. Thanks for the help and the WF.
@Erunai That's greate to hear ! Good job on working this out !
Hey, amazing workflow, thanks for this!
I just have an issue, when I click run it hangs before upscaling saying that some detailer models (bbox and segm) are missing. Do I have to use exactly yours? Even if I try to bypass the detailer step, it won't go beyond the first preview. Any ideas?
You don't need to use mine specificly, but you need to change the model to yours in the detailer model section
@Mittoshura that's what I did. But for an unknown reason the workflow hangs and won't go beyond the first preview. I'm a bit clueless. Do they have to be segm or bbox? Anyway, thanks for your answer! 👍
@SassyAI It really depends on the file type you’re working with.
If it’s a .BBOX file, open the corresponding Detailer node and load the BBOX track into it — this defines the bounding box trail for object tracking or refinement.
If it’s a .SEGM file, you can input it directly into the Detailer as well. In some cases, you can also combine it with the BBOX trail for more precise segmentation and tracking.
@Mittoshura thanks a lot mate, I'll give it a go
amazing workflow, this is also how i found out there soemthing call lora manger, its crazy good. however, when i ran 1.5 for the first time today, the run ended right at the preview node and then stopped, i tried multiple times and its stil lthe same thing
it gave me some weird images with blue skin and green hue all of it looks weird
Maybe it's the loras that you use ? :/ I don't really have an answer for your bug, probably come from your checkpoint or smthg
i ended up just deleteing all the custom nodes i already have and let the manager install based on your workflow and its working. just a quick question though how do i use embedding with your wokrflow?
@ouroboroslapse130 You call them by putting "embedding::thenameofyouembedding" something like "embedding:negative_hand-neg"
@Mittoshura ure awesome
Thank you very much for sharing this excellent workflow!
I have one more question, if you don’t mind.
Is there any way to disable the image filter alert sound in ComfyUI?
I have no idea sadly, probably in the option of comfyui or just mute in windows mixer
is there a way to implement ipadaptor or better controlnet node? The sampled image from the controlnet option always gets cropped at the end
The current node is configured for OpenPose. If you replace both the input image and the model with a more suitable ControlNet, it should function correctly — at least in theory.
I got the image from high res but facedetailer isn't executed
You're going to tell me more than that if you want help
@Mittoshura Color Fix is the last node which is executed and then the workflow stops. Face Detailer and Eyes Detailer are not reached and I don't get the final image
yeah im getting the same issue here, for some reason the face detailer stopped working and im getting a "
0: 384x640 1 hand, 4.9ms
Speed: 1.1ms preprocess, 4.9ms inference, 0.8ms postprocess per image at shape (1, 3, 384, 640)
0: 384x640 1 hand, 4.5ms
Speed: 1.2ms preprocess, 4.5ms inference, 1.7ms postprocess per image at shape (1, 3, 384, 640)
!!! Exception during processing !!! 'NoneType' object has no attribute 'data'
Traceback (most recent call last):
File "G:\ComfyUI_text_to_imag\ComfyUI\execution.py", line 496, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "G:\ComfyUI_text_to_imag\ComfyUI\execution.py", line 315, in get_output_data
return_values = await asyncmap_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "G:\ComfyUI_text_to_imag\ComfyUI\custom_nodes\comfyui-lora-manager\py\metadata_collector\metadata_hook.py", line 165, in async_map_node_over_list_with_metadata
results = await original_map_node_over_list(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "G:\ComfyUI_text_to_imag\ComfyUI\execution.py", line 289, in asyncmap_node_over_list
await process_inputs(input_dict, i)
File "G:\ComfyUI_text_to_imag\ComfyUI\execution.py", line 277, in process_inputs
result = f(**inputs)
^^^^^^^^^^^
File "G:\ComfyUI_text_to_imag\ComfyUI\custom_nodes\comfyui-impact-pack\modules\impact\impact_pack.py", line 876, in doit
enhanced_img, cropped_enhanced, cropped_enhanced_alpha, mask, cnet_pil_list = FaceDetailer.enhance_face(
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "G:\ComfyUI_text_to_imag\ComfyUI\custom_nodes\comfyui-impact-pack\modules\impact\impact_pack.py", line 819, in enhance_face
segm_segs = segm_detector.detect(image, bbox_threshold, bbox_dilation, bbox_crop_factor, drop_size)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "G:\ComfyUI_text_to_imag\ComfyUI\custom_nodes\comfyui-impact-subpack\modules\subcore.py", line 479, in detect
detected_results = inference_segm(self.bbox_model, utils.tensor2pil(image), threshold)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "G:\ComfyUI_text_to_imag\ComfyUI\custom_nodes\comfyui-impact-subpack\modules\subcore.py", line 373, in inference_segm
segms = pred[0].masks.data.cpu().numpy()
^^^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'data'
"
error
What to do in this situation?
Missing Node Types
When loading the graph, following node types were not found
Some nodes require a newer version of ComfyUl (current: 0.3 66). Please update to use all nodes.
Requires ComtyUl 0.3.51:
5aa08a6c-3415-44da-9a51-4d6546548018
194db35f-c3a9-46ee-8b97-29754c018db6
This error means your ComfyUI version is older than what the workflow was built for — some nodes were made with a newer version (in your case, it says they require ComfyUI 0.3.51 or newer).
@Mittoshura Cool cool, it confused me cuz it says I have 3.6 and the required ver is 3.51
@Mittoshura Another thing, The workflow stops at High res img, the detailers don't work, I'm not sure if i'm missing something or what cuz I have all the files and everything
@xnull87 Are you sure you’re using the same models mentioned in the description? Try to disable the detailer and see if it work, i have honestly no idea why it's not working
@Mittoshura I Think I found a flaw in you're V1.5, There's a hidden duplicate node under "Ultralytics Face". Which causes it to not engage the Detailer part of the workflow.
Also, There's no SAM Loader connected to the detailers.
I want to switch from Fooocus to CompUI. do I understand correctly that in order for the template to work, I need to download all the Details Models and Nodes in the description? where should the Detailers and Nodes be stored?
The nodes goes in your "Custom nodes" folder in your comfyui roots, an your detailers models go in "Comfyui/models/ultralytics" !
So I feel kind of dumb, but I can't for the life of me figure out how to get the image to upscale in v1.5. It just shows the initial Raw Image, but never makes it to High Res or Final. Can someone explain how that's supposed to work?
You can add me on discord if you want ! I'll gladly help you ! (my discord : mittoshura)