Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > ComfyUI-Pixaroma

ComfyUI Extension: ComfyUI-Pixaroma

Repo Name

ComfyUI-Pixaroma

Author
pixaroma (Account age: 470 days)
Nodes
View all nodes(1)
Latest Updated
2026-11-06
Github Stars
0.48K

How to Install ComfyUI-Pixaroma

Install this extension via the ComfyUI Manager by searching for ComfyUI-Pixaroma
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter ComfyUI-Pixaroma in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

ComfyUI-Pixaroma Description

ComfyUI-Pixaroma adds nodes and canvas tools for editing, previewing, comparing, and saving images, video, audio, text, and 3D scenes, while managing prompts and organizing complex ComfyUI workflows.

ComfyUI-Pixaroma Introduction

Pixaroma logo

ComfyUI-Pixaroma adds creative tools to ComfyUI workflows. It includes nodes for working with images, video, audio, prompts, text, and 3D scenes, plus canvas tools that make larger workflows easier to navigate. A node is a step in a workflow: it receives an image, a setting, or another value, does a job, and can pass the result to the next step.

You can use Pixaroma to preview and compare images before saving them, resize pictures without guessing pixel dimensions, try several prompts in one session, or label the parts of a complex workflow. Its editors also let you paint, compose images, sketch requested changes, and build 3D scenes inside ComfyUI.

You do not need every tool to benefit from the extension. Start with the task you want to accomplish, then add the relevant nodes. To find them, double-click an empty area of the ComfyUI canvas and search for Pixaroma, or open the 👑 Pixaroma folder in the Node Library.

How ComfyUI-Pixaroma Works

Think of a workflow as a route through a studio. An image-loading node brings in your source picture; a resize or editing node prepares it; a preview node lets you inspect it; and a save node writes the finished file. The wires between nodes carry the image or other information from one step to another.

Pixaroma adds tools for several parts of that route:

  • Prepare an input: Load an image, video, sound file, 3D model, or folder of files. Resize or crop an image to suit the next step.
  • Guide a result: Write a prompt, provide reference images, choose a seed or image size, or turn a section of the workflow on and off.
  • Inspect and finish: Pause to review an image or text, compare alternatives, add an overlay or watermark, and save the result.

Some Pixaroma tools do not process an image at all. Label and Note help explain a workflow to yourself or someone else. Switch, Control Panel, and Set and Get help manage choices and settings across a busy canvas. The toolbar’s Help window and Workflows panel are canvas tools, so you can use them without adding a node.

For example, to explore several compositions, you might connect Resolution Pixaroma to the width and height inputs of your workflow, use Seed Pixaroma to keep or change the starting seed, and view the results with Preview Image Pixaroma. XY Plot Pixaroma can then arrange a comparison of settings in one labeled grid.

ComfyUI-Pixaroma Features

The groups below are organized by creative task. Each linked node name opens its detailed help page, including its inputs and settings.

Images and video

ToolWhat you can do with it
Preview Image (https://workflows.pixaroma.com/help/nodes/preview-image/)View one image or a batch on the node, then save, copy, or open an image.
Load Image Mini (https://workflows.pixaroma.com/help/nodes/load-image-mini/) and Load Image (https://workflows.pixaroma.com/help/nodes/load-image/)Bring a picture into a workflow. Mini keeps the node compact and puts resizing in its gear menu; Load Image offers built-in options such as a longest-side limit, a megapixel limit, fitting inside a size, and cropping to fill.
Image Info (https://workflows.pixaroma.com/help/nodes/image-info/)Get the mask, dimensions, and file name from Load Image Mini’s image information output.
Image Compare (https://workflows.pixaroma.com/help/nodes/image-compare/)Examine two results with a slider, side by side, or in a view that highlights differences.
Load Images from Folder (https://workflows.pixaroma.com/help/nodes/load-images-from-folder/)Process pictures from a chosen folder one by one. You can select fewer pictures or skip ones already completed when working through a large collection.
Image Grid (https://workflows.pixaroma.com/help/nodes/image-grid/)Gather the images from a run into one grid with their file names underneath.
Save Image (https://workflows.pixaroma.com/help/nodes/save-image/)Save to a chosen folder and see the intended file name on the node. A Print resolution option can save PNG and JPG files at 300 ppi.
Remove Background (https://workflows.pixaroma.com/help/nodes/remove-background/)Separate a subject from its background and produce a cutout, a mask, and an inverted mask. Your choice of background-removal model affects which edges it handles best.
Load Video (https://workflows.pixaroma.com/help/nodes/load-video/), Load Video Frame (https://workflows.pixaroma.com/help/nodes/load-video-frame/), and First Last Frame (https://workflows.pixaroma.com/help/nodes/first-last-frame/)Load a clip as frames and sound, pick a particular frame, or extract its opening and closing frames.
Save Mp4 (https://workflows.pixaroma.com/help/nodes/save-mp4/) and Save Video (https://workflows.pixaroma.com/help/nodes/save-video/)Save and preview video. Save Video also supports a chosen folder and file name; its gear settings can use a compatible NVIDIA card for encoding or fall back to the CPU.
Video Tools (https://workflows.pixaroma.com/help/nodes/video-tools/)Join, trim, reverse, loop, and fade up to eight clips with their sound; optionally place one song beneath the whole video.
Reference Images (https://workflows.pixaroma.com/help/nodes/reference-images/)Assign distinct jobs to as many as 12 reference pictures and create a prompt that names them in the form expected by the selected model.
Seamless Tile (https://workflows.pixaroma.com/help/nodes/seamless-tile/) and Seamless Preview (https://workflows.pixaroma.com/help/nodes/seamless-preview/)Create repeating patterns or textures, then inspect them as a 2×2, offset, or half-drop repeat to check how the edges meet.
Save 3D (https://workflows.pixaroma.com/help/nodes/save-3d/) and Load 3D (https://workflows.pixaroma.com/help/nodes/load-3d/)Preview and exchange 3D work. Save 3D can position a model on the ground and save OBJ, GLB, or STL files; Load 3D supplies both the model and an image of it.

Prompts and text

ToolWhat you can do with it
Prompt (https://workflows.pixaroma.com/help/nodes/prompt/)Write prompts using saved @tags, lists, and random choices to reuse ideas without retyping them.
Text (https://workflows.pixaroma.com/help/nodes/text/) and Show Text (https://workflows.pixaroma.com/help/nodes/show-text/)Type text once and send it to several nodes, or display and copy text and values passing through a wire.
Text Join (https://workflows.pixaroma.com/help/nodes/text-join-two/) and Find and Replace (https://workflows.pixaroma.com/help/nodes/find-and-replace/)Combine two, three, or four pieces of text, or change specified words as text moves through the workflow.
AI Prompt (https://workflows.pixaroma.com/help/nodes/ai-prompt/)Use a language model on your own computer with an instruction saved on the node. Presets cover tasks including pattern prompts and prompts for particular image models.
Prompt Each (https://workflows.pixaroma.com/help/nodes/prompt-each/), Prompt Multi (https://workflows.pixaroma.com/help/nodes/prompt-multi/), Prompt From List (https://workflows.pixaroma.com/help/nodes/prompt-from-list/), and Prompt Pack (https://workflows.pixaroma.com/help/nodes/prompt-pack/)Explore multiple ideas: make one picture per prompt in a run, queue separate runs, choose a prompt by number, or paste a block of prompts to queue.
Prompt Stack (https://workflows.pixaroma.com/help/nodes/prompt-stack/)Build a prompt from rows that you can turn on and off. For example, retain the subject while trying different lighting descriptions.
Prompt Reader (https://workflows.pixaroma.com/help/nodes/prompt-reader/)Read prompt information saved inside supported ComfyUI, Automatic1111, or Civitai pictures when that information is present.
Save Text (https://workflows.pixaroma.com/help/nodes/save-text/), Load Texts from Folder (https://workflows.pixaroma.com/help/nodes/load-texts-from-folder/), and Caption Review (https://workflows.pixaroma.com/help/nodes/caption-review/)Save generated text, process a folder of .txt files, or review pictures alongside their captions when preparing a training set. Save Text can also write one named file per entry.
Pause Text (https://workflows.pixaroma.com/help/nodes/pause-text/) and Pause Image (https://workflows.pixaroma.com/help/nodes/pause-image/)Stop a run at a review point so you can inspect or correct text, or inspect an image before a slower step continues.
Video Prompt (https://workflows.pixaroma.com/help/nodes/video-prompt/) and Music Prompt (https://workflows.pixaroma.com/help/nodes/music-prompt/)Prepare model-specific writing: a MiniMax H3 video prompt, or a caption and lyrics for MiniMax Music 3.
Layout Boxes (https://workflows.pixaroma.com/help/nodes/layout-boxes/)Draw where objects should go and describe each one for a layout-guided image prompt. Its <obj_1>-style tags are for FLUX.3; the documented MOSS-Image workflow handles its boxes differently.

Notes, graphics, and editors

Sizes, cropping, and restoration

Workflow controls and specialist tools

Canvas tools

Open the Workflows panel (https://workflows.pixaroma.com/help/nodes/workflows-panel/) from the top toolbar or with Alt+W to find, open, and organize workflow files with picture previews. Open the Help window (https://workflows.pixaroma.com/help/faq/#q-help-window/) from the toolbar or with Alt+H to search the text of Pixaroma’s node help.

Align (https://workflows.pixaroma.com/help/nodes/align/) shows guides as you position nodes. Pixaroma Group (https://workflows.pixaroma.com/help/nodes/pixaroma-group/) adds Run, Mute, Bypass, and Fold controls to a group header. Node Colors (https://workflows.pixaroma.com/help/nodes/node-colors/) lets you recolor nodes and groups, while adaptive node titles (https://workflows.pixaroma.com/help/nodes/adaptive-node-titles/) keep their text readable. Optional Connection FX (https://workflows.pixaroma.com/help/nodes/connection-fx/) highlights compatible slots while you drag a wire, and Run button effects (https://workflows.pixaroma.com/help/nodes/run-button-effects/) change the appearance of ComfyUI’s Run button.

ComfyUI-Pixaroma Models

Pixaroma is a collection of tools, not a single image-generation model. Some nodes work with particular models or formats; selecting one of those nodes does not mean the underlying model is included.

Background removal choices

Remove Background Pixaroma (https://workflows.pixaroma.com/help/nodes/remove-background/) and the AI background-removal buttons in Paint and Image Composer use a BiRefNet model. Choose a file according to the kind of edge you need to preserve:

ModelExpected file nameStated graphics memoryWhen to choose it
Standardbirefnet.safetensors4–6 GBA faster general choice for objects, products, and logos.
High Resolutionbirefnet-hr.safetensors8 GB or moreLarger images with fine edges.
Mattingbirefnet-matting.safetensors8 GB or moreDifficult soft edges such as hair, fur, lace, and fabric.

The files belong in ComfyUI/models/background_removal/. The High Resolution and Matting downloads arrive named model.safetensors; rename each to the exact name shown above so Pixaroma can identify it. See the model-choice FAQ (https://workflows.pixaroma.com/help/faq/#q-bg-models) and renaming FAQ (https://workflows.pixaroma.com/help/faq/#q-bg-rename) if one is not recognized. The Paint and Image Composer editors can also use rembg, but the Remove Background node does not require it; see the rembg FAQ (https://workflows.pixaroma.com/help/faq/#q-rembg-needed).

Other model-specific tools

  • Seamless Tile supports the listed Krea 2, Z-Image, FLUX.2 Klein, FLUX 3 Klein, SDXL, and SD 1.5 workflows for repeating patterns and textures. Seamless Preview helps you check the resulting repeat.
  • Reference Images formats named image references for the supported Qwen Image 2.1, FLUX 3, and MiniMax H3 uses described in its help.
  • Krea2 Enhancer and Krea LoRA Converter serve Krea 2 workflows.
  • H3 Latent Upscale, H3 Audio Sync, H3 Timeline, H3 Long Video, and Video Prompt serve MiniMax H3 tasks. Music Prompt writes for MiniMax Music 3.

Use each linked node page to check its particular inputs before connecting it to a model workflow.

What's New with ComfyUI-Pixaroma

The supplied changelog covers October 4 through November 16, 2026. This is a guided account of the changes shown in that excerpt; it does not imply that earlier or omitted releases are covered.

Version and dateChanges that matter for your work
v1.4.256 · November 16Added Seamless Tile and Seamless Preview for making and checking repeating artwork. AI Prompt gained four pattern-prompt presets, and Save Image gained a 300 ppi Print resolution option for PNG and JPG.
v1.4.254–255 · November 13Improved alignment guides when resizing Nodes 2.0 nodes and corrected Classic resizing behavior. Group Switch more reliably handles collapsed nodes; save-file counters look only in the chosen save folder.
v1.4.253 · November 11Resolution gained a Big size range of roughly 4–17 megapixels, and custom width or height limits rose to 8192. Its ratio and megapixel display became easier to read.
v1.4.252 · November 10Image and video file names can include %run_time%. Cross-platform model-file lookup improved for LoRA Loader and the prompt-writing nodes; Inpaint Crop’s displayed crop now better matches the actual crop.
v1.4.249–251 · November 5–6Sizes gained quick labels such as HD and 4K. Save Video gained compatible NVIDIA encoding and a bitrate target; saved video color information improved. Reference Images gained P-H3 Image detection, and Firefox text-box scrolling was fixed.
v1.4.244–248 · October 31–November 2Text-box scrolling was improved; Group Switch gained more control over nested groups. Fixes addressed pasting nodes, opening workflows containing Pixaroma subgraphs, group switching, and running selected setting nodes from scripts without a browser.
v1.4.238–243 · October 28–30Added H3 Latent Upscale. Load Video became more reliable with some phone videos; H3 Audio Sync preserved the supplied song’s quality. Fixes also covered Number Pick connections, Group Switch performance, and clearer Duration and H3 Long Video guidance.
v1.4.230–237 · October 23–27Added Reference Images, Video Tools, and Krea2 Enhancer. Video Tools improved audio levels through crossfades; Load Video better handles oversized clips; several nodes improved their behavior inside subgraphs.
v1.4.227–229 · October 21–22Added H3 Timeline and H3 Long Video for longer continuous video work. AI Prompt gained MOSS-Image presets; prompt-writing nodes gained a fallback for some graphics cards. Text and prompt nodes gained more flexible text sizing.
v1.4.216–226 · October 16–20Added Layout Boxes and expanded Resolution and Sizes with a batch setting. Updates also improved upright phone-video handling, workflow controls, model-list refreshes, prompt presets, canvas dialogs, and compatibility with newer ComfyUI versions.
v1.4.210–215 · October 12–15Resolution and Sizes gained a latent output. Image previews and other result displays became more reliable when switching workflow tabs. Keyboard shortcuts, video controls, image masks, and several editor interactions were corrected.
v1.4.203–209 · October 8–11Added Image Grid, Text Match, Load Texts from Folder, and Caption Review. Folder processing gained path output and ways to skip completed pictures; Save Text gained one-file-per-entry output. The Help window gained a Latest version button.
v1.4.193–202 · October 4–6Improved generation information saved for Civitai, prompt reading from JPG and WebP, video generation information, transparent text overlays, and compatibility with newer ComfyUI behavior. The supplied excerpt ends partway through the October 4 entry, so no further details of that entry are included here.

Troubleshooting ComfyUI-Pixaroma

A node is empty, or its buttons disappear after an update

  1. Restart ComfyUI.
  2. Hard-refresh the browser with Ctrl+Shift+R; on a Mac, use Cmd+Shift+R.
  3. Reopen the workflow and check the node again. The browser may have retained an older version of Pixaroma’s interface. Follow the illustrated refresh guide (https://workflows.pixaroma.com/help/refresh/) if the problem remains.

A workflow shows red “missing nodes”

  1. Open the missing-nodes guide (https://workflows.pixaroma.com/help/missing-nodes/) and identify the node names shown in red.
  2. Check whether the workflow requires Pixaroma or another node pack.
  3. After correcting the missing-node issue, restart ComfyUI and hard-refresh the browser before reopening the workflow.

A workflow says a model is missing

  1. Note the model or file name named by the workflow.
  2. Use the missing-models guide (https://workflows.pixaroma.com/help/missing-models/) to identify the file it requires.
  3. For Pixaroma background removal, compare the file name and location with the Background removal choices table above. In particular, rename High Resolution or Matting downloads that still have the generic name model.safetensors.
  4. Restart ComfyUI and try the workflow again.

A large video stops before processing

Load Video may stop when a long or 4K clip exceeds the memory available to the computer. Read the Max frames value shown by the node, then try fewer frames or a smaller source. This prevents a run from remaining frozen while attempting to load more than fits.

A video duration value keeps snapping back

Open Duration Pixaroma’s gear settings and raise MAX if the duration you entered is outside its current range. The node displays its accepted range and explains why an out-of-range value was reset.

A prompt-writing node fails on the graphics card

If AI Prompt, Music Prompt, or Video Prompt encounters the documented CUDA problem with a Qwen3.5 model, open the node’s gear settings and turn off Faster writing (MTP), then run it again. For other errors, consult that node’s help page before changing unrelated workflow settings.

Frequently asked questions

Where are the Pixaroma nodes? Double-click empty canvas space and search Pixaroma, or browse the 👑 Pixaroma folder in the Node Library.

How do I check which version I have? Open the Help window with the top-toolbar ? button or Alt+H; the version appears at its bottom right. You can also add Version Check Pixaroma (https://workflows.pixaroma.com/help/nodes/version-check/).

Why can I see a node but not its latest controls? Restart ComfyUI and hard-refresh the browser after an update. An older cached interface can leave controls missing or out of date.

Which background-removal model should I choose? Start with Standard for general objects. Consider High Resolution for large images with detailed outlines, or Matting for soft, intricate boundaries such as hair. The latter two list higher graphics-memory requirements.

Do I need rembg to use Remove Background Pixaroma? No. The node uses the listed BiRefNet models. The Paint and Image Composer editors can also use rembg; the FAQ explains the distinction (https://workflows.pixaroma.com/help/faq/#q-rembg-needed).

Can I make many variations without rebuilding my workflow? Yes. Use Prompt Each or Prompt Multi for prompt variations, Resolution or Sizes for image dimensions and batch counts, and XY Plot to compare settings in a labeled grid.

Learn More about ComfyUI-Pixaroma

ComfyUI-Pixaroma Related Nodes

RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

ComfyUI-Pixaroma detailed guide | ComfyUI