ComfyUI Level Pixel
Main nodes of the Level Pixel company (aka levelpixel, LP). Includes convenient nodes for working with images from folders; counting files in a folder; cleaning memory;…
Nodes (92)
The forgiving stringify-anything node
Turning a checkbox into 0 or 1
Turning a toggle into readable text
Turning a mask into width/height numbers
Prompt in your language, encode in English
Flattening a dropdown's picks into a string
Serializing a JSON object back to text
Parsing text into a real JSON object
Counting whatever's connected to it, with a running offset
The default sizing mode for Inpaint Crop
Pinning Inpaint Crop to an exact resolution
Sizing Inpaint Crop relative to the mask, not a target resolution
Keeping Inpaint Crop within a min/max window
Pausing the graph for a fixed number of seconds
Extending Inpaint Crop's area asymmetrically per side
Generating a transparency-test checkerboard
Counting matching files in a folder
Looking up a value by key, without crashing if it's missing
Rounding a float down, up, or to nearest
Turning a number into text
Resolving a real filesystem path from ComfyUI's model registry
Resolving output/input/temp for the HTTP API
Stepping through a folder one file at a time
Get Iterator Data From Image Folders [LP] — ComfyUI Node Guide
Get Iterator Data From Videos [LP] — ComfyUI Node Guide
Get JSON Value [LP] — ComfyUI Node Guide
Hard Unload All Models [LP] — ComfyUI Node Guide
Simple Float Slider - Hundredths Step [LP] — ComfyUI Node Guide
Image Data Iterator [LP] — ComfyUI Node Guide
Image Loader From Path [LP] — ComfyUI Node Guide
Image Overlay [LP] — ComfyUI Node Guide
The crop half of the crop-edit-stitch inpainting pattern
Pasting the edited fragment back without touching anything else
Convert Int To Bool [LP] — ComfyUI Node Guide
Convert Int To Float [LP] — ComfyUI Node Guide
Bridging ComfyUI's strict socket types
Iterator [LP] — ComfyUI Node Guide
Check JSON Key [LP] — ComfyUI Node Guide
Get JSON Length [LP] — ComfyUI Node Guide
Stringify JSON [LP] — ComfyUI Node Guide
Keep Only English Tags [LP] — ComfyUI Node Guide
Keep Only English Words [LP] — ComfyUI Node Guide
Load Image [LP] — ComfyUI Node Guide
Load LoRA Tag [LP] — ComfyUI Node Guide
Merge JSON Node [LP] — ComfyUI Node Guide
Unload Model [LP] — ComfyUI Node Guide
Set one key on a JSON object, then chain the next
Force the text encoder onto CPU to free up VRAM
Freeing VRAM from the reference-image encoder
Dodging the out-of-memory crash right at decode
Turn raw JSON text into an object you can actually wire up
Bundle a dozen wires into one before they turn your graph to spaghetti
The one node that's both ends of the wire at once
Unbundle a pipe back into the individual wires you actually need
A preview that only fires when something's actually watching
Instruction-based photo editing without loading the model yourself
Strip a blocklist out of a comma-separated tag string
The same filter, built for prose instead of a tag list
Cleaning up an assembled tag string
Keeping an image and its masks in perfect sync
A quick resize with three fixed presets and four algorithms
Reassembling filtered tags into one ordered prompt
One number, wired everywhere you need it
A text preview that skips itself when nothing's watching
Always runs, always shows you the string, connected or not
A 0-to-1 drag bar for tweaking values live
An end-of-queue memory sweep
A lighter-touch memory sweep mid-workflow
A text-in, text-out utility whose exact rule isn't written down anywhere
Step through a list of strings across queued runs
A single-line text box, for when a full Text node is overkill
Turning text into a real boolean
Sneaking a wired string into a dropdown-only input
Parsing text into a real numeric value
Turn text into a whole number, no rounding options attached
Get both the float and a rounded int from one string, one node
Keep or drop whole categories of tags in one pass
Pull out just the tag categories you actually want
Figure out what category each of your tags actually belongs to
Drop whole categories of tags, keep everything else
Combine two tag strings into one without duplicating everything
Cutting specific tags out of a prompt string
Fuzzy find-and-replace for tag strings, not exact-match
Pick which image flows through based on which tags are present
A 0.0–1.0 dial that snaps to clean values
The old A1111 {this|that} wildcard trick, as a node
A plain multi-line text box for feeding strings into other nodes
Up to nine find-and-replace pairs in one node
Turn up to ten text boxes into one string or a real list
Auto-translating your prompt to English, free
Free Google/MyMemory translation, full language list
Free translation with a short, human-readable language list
Level Pixel nodes for ComfyUI
The purpose of this package is to collect the most necessary and atomic nodes for working with any tasks, adapted for use in cycles and conditions. The package of nodes is aimed at those users who need all the basic things to create multitasking complex workflows using multimodal neural models and software solutions.
In this Level Pixel node pack you will find: Inpaint Crop and Stitch, Pipeline, Tag Category Filter nodes, Model Unloader nodes, File Counter, Object Counter, Image Loader From Path, Load Image, Fast Checker Pattern, Float Slider, Load LoRA Tag, Image Overlay, Conversion nodes.
Recommend that you install the advanced node package from Level Pixel Advanced for Multimodal Generators, Qwen2.5-VL gguf, LLM, VLM, RAM, Autotaggers, RemBG nodes:
https://github.com/LevelPixel/ComfyUI-LevelPixel-Advanced
The official repository of the current node package is located at this link:
https://github.com/LevelPixel/ComfyUI-LevelPixel
Like our nodes? Then we'd be happy to see your star on our GitHub repository!
Contacts, services and support
Our official Patreon page:
https://www.patreon.com/LevelPixel
On Patreon you can get services from us on issues related to ComfyUI, Forge, programming and AI tools.
You can also support our project and support the development of our node packages.
For cooperation, suggestions and ideas you can write to email:
[email protected]
Installation
Installation package using ComfyUI Manager (recommended)
Install ComfyUI Manager and do steps introduced there to install this repo 'ComfyUI-LevelPixel'.
The nodes of the current package will be updated automatically when you click "Update ALL" in ComfyUI Manager.
Alternative installation package
Clone the repository:
git clone https://github.com/LevelPixel/ComfyUI-LevelPixel.git
to your ComfyUI custom_nodes directory
The script will then automatically install all custom scripts and nodes.
It will attempt to use symlinks and junctions to prevent having to copy files and keep them up to date.
- For uninstallation:
- Delete the cloned repo in
custom_nodes - Ensure
web/extensions/levelpixelhas also been removed
- Delete the cloned repo in
- For manual update:
- Navigate to the cloned repo e.g.
custom_nodes/ComfyUI-LevelPixel git pull
- Navigate to the cloned repo e.g.
Features
All nodes Level Pixel in this package:
Inpaint Crop and Stitch
Inpaint Crop and Inpaint Stitch nodes allow you to work effectively with image fragments.
Inpaint Crop cuts out a part of the image using a given mask, increases/decreases the cut out image fragment to the target size, and then, after any third-party modifications to the image fragment (regeneration, manual correction, repainting, etc.), using Inpaint Stitch, it stitches this modified fragment back into the main image (after having reduced/increased it to the size it had when cutting the fragment out of the main image).
This is incredibly convenient for Inpaint, and especially convenient for automation.
In addition, the peculiarity of these Inpaint Crop and Stitch nodes is that they have a special pin that accepts different parameters for the specified size of the cut out fragment. One of the key ones is "Cropped Aspect Size Parameters" (or the "aspect size" value embedded in Inpaint Crop in the mode widget), which allows you to set the target resolution for the cut fragment, and the program itself calculates the required dimensions, which can be incredibly useful for automation and regeneration of a picture fragment.
In addition, Inpaint Crop and Stitch can easily work with Batch images, masks and context masks. This can be very useful in many scenarios when you need to mass-cut images by mask, and then also mass-return them to the main images. The only limitation is that the target size of the masks must be the same for batch images.
The image below shows an example of using the batch mode.
Resize Image and Masks
Resize Image and Masks is a node that allows you to do a great thing - bring an image/mask/context mask or a batch of images/masks/context masks to one target size.
This can be incredibly useful when you need to bring images to one target size along with their masks, without any errors and in one go.
In addition, you can also use this node for special cases of adjusting the size of an image (or mask, or context mask) to the target.
An example of how you can use this node is below (and also on the previous image in Inpaint Crop and Stitch).
The core functionality is taken from ComfyUI-Inpaint-CropAndStitch and belongs to its authors.
Tag Category Filter nodes
A set of nodes that allow you to filter tags by category. There is an option to remove or leave certain categories of tags, there is a function for defining categories of all tags, there is a function for removing certain tags.
Nodes are very convenient because you can use them to remove unnecessary tags by certain categories, for example, to clean up tags and prepare them for use. You can use this to get certain prompts from an image (for example, if you need a description of only the background from an image - you can get this category of tags if you set the "background" category in Tag Category Keeper).
The core functionality is taken from comfyui_tag_fillter and belongs to its authors.
Load LoRA Tag
LoRA loader from text in the style of Automatic1111 and Forge WebUI. For this version of loader, text output for errors when loading LoRA has been added as widget on node.
The core functionality is taken from comfyui_lora_tag_loader and belongs to its authors.
Model Unloader nodes
A node that automatically unloads all models from memory. It must be added to a sequential chain of nodes in the workflow. There are three versions of this node: Hard (complete unloading of all models from memory), Soft (without unloading models from memory, just soft cleaning of memory from garbage), Model Unload (unloading of specifiy model from memory), and "Soft Full Clean RAM and VRAM" for the case when it is necessary to unload all models from memory and clean memory from garbage after the workflow is completed (the task in the queue is finished).
File Counter
A simple counter of files in a given folder. Convenient when you need to count the number of files in a certain format (for example, for subsequent use in loops or conditions).
Image Loader From Path
Loads images from a specific folder or path. It is convenient because you can specify both absolute paths and local paths (from the input folder), as well as the ability to sequentially receive images into the workflow at each step of the cycle by number. In addition, you can load images in batches with a certain number of loaded images in a batch. And the cherry on the cake - you can get a list of image file names at the output.
Load Image
This is a new image loading node that can retrieve the name of the files you load into your workflow.
Image Overlay
A node that allows you to overlay one image on another with the ability to specify a mask. In this package, Image Overlay has an extended range of specified sizes for the final image, and also has another standard image size.
The core functionality is taken from efficiency-nodes-comfyui and belongs to its authors.
Fast Checker Pattern
Quickly creates a background image with a checkerboard pattern according to the specified parameters for subsequent testing of images with a transparent background. You need to combine the resulting background image with your image with a transparent background in other ComfyUI nodes (at the moment there is no universal node, but perhaps we will make one in the future).
Simple Float Slider
Simple Float Slider is a handy slider from 0.0 to 1.0 to conveniently manage variables in your workflow. The min and max values cannot be changed on the interface (but you can change these values inside Python if you really need to).
The pack contains two additional sliders - "Simple Float Slider - Tenths Step" and "Simple Float Slider - Hundredths Step" for working with more precisely defined values in tenths and hundredths (work correctly only if you have not changed the value of "Float widget rounding decimal places" in the ComfyUI settings. If you have changed it, then return the value to 0).
The core functionality is taken from comfyui-mixlab-nodes and belongs to its authors.
Other nodes
There are a few more nodes in this package that have some unusual uses:
- Google Translate
- Count Objects - counts the number of objects that were fed to the input (this can be a list or one single object). Accepts any type of input.
- Preview Image Bridge - only output an image to the screen if there is a connection to the output node. Useful in loops and conditions where the execution of this node is not required due to current conditions (variables).
- Show Text Bridge - only output text to the screen if there is a connection to the output node. Useful in loops and conditions where the execution of this node is not required due to current conditions (variables).
- Show Text - output text to the screen with mandatory execution. The node is executed in any case, whether the output is connected or not.
- Text - a simple node for entering multi-line text (similar to Prompt from other node packages).
- String - a simple node for entering single-line text (similar to String from other node packages).
- Conversion nodes - a variety of different nodes that allow you to transform different types of variables into other variables. The big difference from other current node packages is that they cover a larger number of variable types. Conversion nodes: StringToFloat, StringToInt, StringToBool, StringToNumber, StringToCombo, IntToString, FloatToString, BoolToString, FloatToInt, IntToFloat, IntToBool, BoolToInt.
- Pipe - extremely useful and extremely easy to use node for building a beautiful pipeline. One Pipe node is both an input and an output, so I recommend using it where it is absolutely necessary. In addition, there are standard Pipe In and Pipe Out, if you want aesthetics.
- JSON - a node for working with JSON data.
About LLM, LLaVa, VLM, Autotagger, RAM nodes
All LLM nodes have been moved to a separate ComfyUI Level Pixel Advanced node package, as such nodes require the skill of configuring programs, drivers and libraries for correct use, as well as due to constant changes and other frequent changes that may affect all other functionality of the current node package. In addition, some technologies based on neural networks tend to quickly become obsolete, so they will be in a separate ComfyUI Level Pixel Advanced package and then ComfyUI Level Pixel Archived package.
Link to Level Pixel Advanced nodes with LLM nodes:
https://github.com/LevelPixel/ComfyUI-LevelPixel-Advanced
Link to Level Pixel Archived nodes with obsolete nodes:
https://github.com/LevelPixel/ComfyUI-LevelPixel-Archived
Update History
v1.3.4 - 24-02-2026 - Added JSON nodes and Qwen Image Edit API nodes
v1.3.0 - 08-06-2025 - Added powerful Inpaint Crop and Inpaint Stitch nodes, as well as Resize Image and Masks
v1.2.0 - 27-05-2025 - The node package is divided into two independent packages - a package with logical nodes ComfyUI-LevelPixel and a package with wrapper-nodes for neural models ComfyUI-LevelPixel-Advanced
The license for this package has been changed from Apache 2.0 to GNU GPLv3
Credits
ComfyUI/ComfyUI - A powerful and modular stable diffusion GUI.
Tag Filter nodes for ComfyUI/comfyui_tag_fillter - Best tag filter by category nodes for ComfyUI.
Load LoRA Tag node for ComfyUI/comfyui_lora_tag_loader - Thanks to the author for this great node for LoRAs!
Efficiency-nodes-comfyui/efficiency-nodes-comfyui - Thanks for Image Overlay!
ComfyUI-Inpaint-CropAndStitch/ComfyUI-Inpaint-CropAndStitch - Thanks for Inpaint Crop and Inpaint Stitch!
Comfyui-Simple-Json-Node/Comfyui-Simple-Json-Node - Thanks for some JSON nodes!
License
Copyright (c) 2024-present Level Pixel
Licensed under GNU GPLv3