ComfyUI-easygoing-nodes Introduction
<div align="center">
<img width="800" height="343" alt="ComfyUI Easygoing Nodes banner image" src="Images/easygoing_banner_image.png">
ComfyUI-easygoing-nodes is a collection of custom nodes and workflow tools for ComfyUI. It helps you with practical tasks around image creation: loading and identifying models, keeping generated images organized, saving prompts alongside images, managing memory use, adjusting color, and comparing results.
You might use it to answer questions such as “Which model made this image?”, “What changed when I used a different upscaler?”, or “Can I free memory after encoding my prompt?” The tools are independent, so you can add the ones you need to an existing workflow.
The extension supports ComfyUI Nodes 2.0 and the V3 node schema as of v2.0.0. One exception is Workflow Image screenshot export, which currently works only in Nodes 1.0.
How ComfyUI-easygoing-nodes Works
A ComfyUI workflow passes information between connected nodes. ComfyUI-easygoing-nodes adds nodes that work at different points in that process:
- Before generation, a model-loading node can provide a cleaner version of a model’s filename for naming your outputs.
- While processing a prompt, the CLIP text-encoding node can release the text encoder from video memory after it finishes its job.
- After generation, you can save prompt details in a PNG, adjust an image’s color and tone, or compare two images to see what changed.
Think of these as tools around your image-making process rather than a new image-generation model. For example, if you generate the same scene with two upscalers, Image Difference Checker can show where their outputs differ. If you return to a finished PNG later, Save Image With Prompt can help you find the prompt information without opening the entire workflow.
ComfyUI-easygoing-nodes Features
Workflow Screenshots
<div align="center">
<img width="800" height="535" alt="Workflow Image screenshot example" src="Images/workflow_image_screenshot_with_box.png">
Workflow Image exports a high-resolution PNG of your current workflow with embedded metadata. It is useful when you want a readable visual record of how nodes were connected.
In Nodes 1.0, right-click the canvas, then choose Workflow Image → Export PNG. The export uses 4× resolution, which helps node labels and connections remain clear when viewed on high-density screens or enlarged.
Important: Screenshot export is not yet supported in Nodes 2.0. Switch to Nodes 1.0 to use this feature. The implementation builds on work from ComfyUI-Custom-Scripts and includes fixes for more recent ComfyUI frontend changes.
Model Loading and Cleaner Output Names
<div align="center">
<img width="800" height="281" alt="Load model with cleaned name nodes" src="Images/load_model_with_name_clean.png">
Load Checkpoint (with Name Clean) and Load Diffusion Model (with Name Clean) help you load models and obtain a simpler name for each one. They list models from models/checkpoints and models/diffusion_models.
Their model_name_clean output removes common technical labels from filenames, regardless of letter case. These include precision and format markers such as fp16, bf16, fp8, and int8; file extensions such as .safetensors, .pth, and .pt; and trailing hyphens or underscores. Other removed markers include fp32, svd, scaled, mxfp8, nvfp4, convrot, hq, e4m3fn, e4m3, e5m2fn, and e5m2.
The result is easier to read in output names. You can use the cleaned name with a trailing / in a Save Image node’s filename_prefix to organize generated images into a subfolder for each model. These nodes load models you already have; the extension does not introduce a separate set of image-generation models.
CLIP Text Encode (with Offload)
<div align="center">
<img width="320" height="242" alt="CLIP Text Encode with Offload node" src="Images/clip_text_encode_with_offload_node.png">
CLIP Text Encode (with Offload) can take the place of the standard CLIP Text Encode node. It encodes your prompt, then explicitly frees the text encoder from VRAM (your graphics card’s memory). It can optionally free it from RAM (your computer’s main memory) too.
This may help when available memory limits the rest of your workflow. The performance trade-off depends on your hardware and how often the encoder is needed again, so compare runs with and without offloading to see which works better for you.
Save Image With Prompt
<div align="center">
<img width="800" height="556" alt="Save Image With Prompt sample workflow" src="Images/save_image_with_prompt_sample_workflow.png">
Save Image With Prompt stores useful generation details as separate fields in a PNG’s metadata, alongside the full workflow JSON. The fields are positive_prompt, additional_prompt, negative_prompt, caption, and seed.
This makes it easier to inspect and copy prompt information in a metadata-capable image viewer, such as XnView MP (https://www.xnview.com/en/xnview-mp/) or digiKam (https://www.digikam.org/). It is particularly helpful when you revisit an image and want its prompt without searching through a workflow.
Seed tip: Seeds are integers in ComfyUI. Use a Preview as Text node, or an equivalent conversion node, to turn a seed into a string before connecting it to this node.
HDR Effects with LAB Adjust
<div align="center">
<img width="800" height="362" alt="HDR Effects with LAB Adjust sample workflow" src="Images/hdr_lab_sample_workflow.png">
HDR Effects with LAB Adjust offers tone and color controls for a stronger, more deliberate look. It works in LAB color space, which separates perceived brightness from color information. In practical terms, this lets you adjust light and color with more independent control.
You can adjust shadows, highlights, gamma, contrast, color boost, and individual LAB channels. For example, you might change shadows and highlights to shape a high-contrast scene, then refine the color controls to reach the look you want. Experiment with small adjustments and compare the result with the original image.
The HDR processing is based on ComfyUI-SuperBeasts, with additional color controls.
Image Difference Checker
<div align="center">
<img width="800" height="594" alt="Image Difference Checker sample workflow" src="Images/image_difference_checker.png">
Image Difference Checker compares two similar images. It produces a difference map to show changed areas, MAE and SSIM similarity measurements, and tone-curve and color analysis.
Use it when a visual comparison alone is difficult. For instance, you can compare an image before and after upscaling, or check how closely a VAE reconstruction matches its source. The difference map helps you locate changes, while the measurements offer another way to compare results. For a meaningful comparison, use two images that are intended to represent the same content.
After changing the checker’s settings, you can use ComfyUI’s partial execution (the blue play button) to rerun only from this node onward rather than repeating earlier workflow steps.
Model Merging
The extension also includes hierarchical model merging nodes, including Model Scale SDXL, and related utilities for advanced blending while preserving model structure. These are intended for artists who want to experiment with combining models; they are not additional pretrained models included with the extension.
What’s New with ComfyUI-easygoing-nodes
v2.1.7
Image Difference Checker changed its brightness calculation from Rec.601 to Rec.709. The new calculation is used for SSIM and the Brightness Difference map and better matches the sRGB/Rec.709 color space typically used by ComfyUI images.
Comparison note: SSIM values, the brightness difference map, and the Brightness row in the tone table can differ from results produced by earlier versions. Do not treat measurements from before and after this update as directly comparable.
v2.1.2
Image Difference Checker replaced its grayscale difference map with a Brightness Difference map calculated using Rec.601 brightness weights. It also added a gray brightness curve over the RGB histogram in the tone-curve graph.
The output name changed from grayscale_diff_map to brightness_diff_map. If an older saved workflow has a connection to the former output, reconnect that link after updating.
v2.1.0
The Workflow Image export script was updated in response to a ComfyUI Nodes Manager security issue.
v2.0.0
This major update added support for Nodes 2.0 and the V3 node schema. It also introduced automatic migration for nodes in existing workflows, although migration may not succeed in every environment.
The experimental sdxl_clip.py replacement for attention-mask and tokenization behavior was removed because recent ComfyUI core updates made it ineffective. If nodes from an older workflow appear missing or broken after updating, remove those instances and add the new nodes in their place. Updating ComfyUI itself is also recommended when you encounter compatibility issues.
Troubleshooting ComfyUI-easygoing-nodes
Workflow Image export is unavailable
- Check whether you are using Nodes 2.0.
- Switch to Nodes 1.0.
- Right-click the workflow canvas and choose Workflow Image → Export PNG.
Screenshot export currently works only in Nodes 1.0.
Nodes are missing or broken in an older workflow
- Check that your ComfyUI version supports the V3 node API (
comfy_api.latest); older builds without it will not register these nodes. - Update ComfyUI if compatibility appears to be the problem.
- Reopen the workflow to allow the extension’s automatic migration to run.
- If a particular node still does not work, delete that old instance and add the corresponding new node.
A difference-map connection broke after an update
- Find Image Difference Checker in the affected workflow.
- Look for a link that previously used
grayscale_diff_map. - Reconnect it to the renamed
brightness_diff_mapoutput.
This output-name change was introduced in v2.1.2.
Similarity numbers changed after updating
Check which version produced each result. In v2.1.7, the checker changed from Rec.601 to Rec.709 brightness calculations. As a result, older and newer SSIM and brightness-related results should not be compared as though they used the same method.
Frequently Asked Questions
Does this extension include image-generation models?
No separate set of pretrained image-generation models is described here. Its loader nodes help you work with models in ComfyUI’s model folders, and its merging nodes provide tools for blending models.
Will CLIP offloading always make my workflow faster?
Not necessarily. It is intended to help manage limited VRAM or RAM, and its performance impact depends on your hardware. Test both settings in your own workflow.
Can I see saved prompts without reopening ComfyUI?
The Save Image With Prompt node stores prompt fields in PNG metadata. You can inspect them in an image viewer that displays that metadata, such as XnView MP or digiKam.
Can I use Workflow Image export in Nodes 2.0?
Not currently. Use Nodes 1.0 for that feature.
Learn More about ComfyUI-easygoing-nodes
- English guide for v2.0.0 (https://www.ai-image-journey.com/2026/08/comfyui-easygoing-nodes-v2.html) and Japanese guide for v2.0.0 (https://note.com/ai_image_journey/n/n7a61e769edab)
- English guide for v1.0.0 (https://www.ai-image-journey.com/2025/09/comfyui-easygoing-nodes.html) and Japanese guide for v1.0.0 (https://note.com/ai_image_journey/n/n5bb33311b866)
- ComfyUI project, for learning more about the workflow system these tools extend
The author’s extension is available under the MIT License.
