ComfyUI Custom Extensions
Browse 6,398 ComfyUI extensions and 45,165 custom nodes.
🌐 ARMD (Adaptive Regional Mixture of Diffusers) is a region-based diffusion system for ComfyUI, designed for semantic upscaling with per-region conditioning, optional VLM captioning, and adaptive regional reconstruction.
Image Tile Split with Region-Aware Prompting for ComfyUI
Tiled regional prompting + tiled VAE decode with seam-free blending for ComfyUI
String utility nodes for ComfyUI: attach titles to prompts, loop them through your workflow, assemble save paths, and derive a distinct seed per image.
Natural language prompt editing and workflow enhancement tools for ComfyUI.
INT+ node for ComfyUI with control after generate style functionality for signal-controlled increment/decrement/randomize operations.
ComfyUI-Ideogram4-Luna is a custom node pack for building structured Ideogram 4 JSON prompts with a visual canvas editor supporting bounding boxes, regions, colors, text elements, and import/export.
ComfyUI-StylePromptLibrary is a custom node pack for selecting reusable image style descriptions from a 306-entry JSON library with cascading pickers, live search, and style strength controls.
A package implementing a Bitalino device ComfyUI custom node.
ComfyUI custom node collection including face direction detection, image processing, and example nodes. (Description by CC)
Experimental ComfyUI custom node wrapper for ByteVisionLab/DreamLite, providing text-to-image and text-guided image editing through official pipeline API.
Compressed-sensing hybrid 4K image-to-video: RAFT flow transfer, dual-channel Laplacian recompose, CS tile scheduler, keyframe refresh helpers.
Inject up to 6 reference images into conditioning for FLUX.2 (and other edit models) in a single node, each with its own strength (0 = skip). Built-in VAE encode.
Privacy-Protected Saver node for ComfyUI that encrypts and archives output images into password-protected ZIP files or videos for secure storage and privacy protection.
ComfyUI custom nodes for text/string manipulation, LoRA selection, image outpainting, YOLO object detection, and interactive bounding box drawing for object-move workflows.
Nodes: KepRotateImage
A collection of useful user interface extensions for ComfyUI to improve workflow and provide helpful feedback.
Custom math, string, and utility nodes for ComfyUI workflows including LoRA preview and group input management. (Description by CC)
Seed-based prompt line picker and VRAM-aware video frame limiter for ComfyUI
Generate and edit images in ComfyUI with OpenAI GPT Image models. Includes dedicated Generate and Edit nodes, multi-image reference input, size presets, and support for both the official OpenAI API and OpenAI-compatible relay endpoints.
HighSync lip-sync nodes for ComfyUI with native audio processing, returning processed frames and encoded mp4 output.
A custom node package wrapping the official KeySync inference pipeline for ComfyUI, enabling advanced motion tracking and synthesis workflows. (Description by CC)
Gemini native generateContent custom nodes for ComfyUI with Nano Banana, Nano Banana 2, and Nano Banana Pro, supporting both Google direct access and Gemini-compatible native relay endpoints.
A custom node for ComfyUI designed to load, organize, and preview text prompts from local .txt files with support for multiline prompts and real-time dynamic previews.
A ComfyUI wrapper implementation of ThinkSound - an advanced AI model for generating high-quality audio from text descriptions and video content using Chain-of-Thought reasoning.
A simple node for loading DDS (DirectDraw Surface) texture files into a PyTorch tensor. The node is designed to be used in visual programming frameworks that support custom nodes (e.g., ComfyUI, AUTOMATIC1111).
A ComfyUI custom node for recutting a frame sequence onto a safe processing canvas for outpainting, image/video generation, and model-specific size constraints.
Utility nodes for ComfyUI by Cornman
Liquid Postman (Gemini Image) - ComfyUI node for Gemini image generation
Auto face detection + inpainting node using SAM3
A custom node extension that adds project management functionality to ComfyUI.
connection nodes for api requests, fully supports promptLAB
AI Horde integration as custom node(s) for ComfyUI
A production-grade collection of specialized ComfyUI nodes for the SiliconFlow Image Generation API.
Run Tencent's state-of-the-art Hy-MT2 multilingual translation models directly inside ComfyUI with no HuggingFace cache pollution, models live cleanly inside your ComfyUI directory.
A ComfyUI custom node package for running state-of-the-art open-source vision-language models locally.
Crop and paste nodes with metadata for detail stitching in ComfyUI
A small ComfyUI custom node that draws bounding boxes from DeepSeek OCR output, crops each bbox region, and can paste processed crops back to the original coordinates.
A collection of logic and comparison nodes for ComfyUI. This package adds building blocks for boolean logic, arithmetic comparisons, and string comparisons, making it easier to design conditional workflows directly in ComfyUI.
ComfyUI custom node that patches loaded models (Flux, SDXL, etc.) to run Linear layers in FP8 or MXFP8, unlocking tensor-core throughput on Ada Lovelace (RTX 4090) and Blackwell (RTX 5090) GPUs.
A Matchering audio mastering/matching node for ComfyUI.
ComfyUI custom nodes for NAFNet image restoration (denoising, deblurring, stereo SR)
This repository provides ComfyUI custom nodes that allow ComfyUI workflows to call an external LCM Stable Diffusion server running on Rockchip NPUs (RK3588 / RKNPU2) over HTTP.
An extension to modify the browser tab title when running ComfyUI workflows.
Metadata for loaded models
A lightweight ComfyUI extension node that runs OpenAI Whisper to transcribe up to 3 audio inputs, merges and sorts timestamped segments, and outputs a single timestamped transcript string with speaker labels.
Nodes to switch image input or output path with boolean conditions
Converts a number (e.g., 2 or 15) into written text (e.g., two or fifteen). Works for english and german up to numbers as high as 9999 without problems.
ComfyUI workflow scanner that cross-scans workflows to detect unused models and custom nodes with high precision using resolver-based analysis. (Description by CC)
A growing collection of custom ComfyUI nodes built while learning AI image generation — practical tools and helpers picked up along the way.
A custom node for ComfyUI that receives a comma-separated string and displays each item as a draggable button that can be toggled on or off.
This package provides custom nodes to ComfyUI to POST data to a secure API.
Import ComfyUI workflows from images with embedded metadata. Adds a toolbar Import button, drag-and-drop dialog and Ctrl+Shift+I shortcut. Supports legacy and new UIs, common image formats, opens each image in a new workflow tab.
ComfyUI plugin providing a toolbox of features, including automatic generation of double eyelid position masks. (Description by CC)
Comprehensive set of ComfyUI custom nodes for interacting with a Sora-compatible REST API, supporting text-to-video, image-to-video, and video-to-video generation.
Custom node for ComfyUI that collects key workflow settings in one place and exposes them as individual outputs.
A small collection of custom nodes for ComfyUI, featuring Max Quick Image Size node for image-to-image or image-to-video workflows. (Description by CC)
A custom node for ComfyUI that allows you to send an image + a message (e.g. your prompt) to a Discord channel via a webhook.
One-click model switching for ComfyUI. Auto-configures CLIP, VAE, resolution, prompts, seed, and sampler settings. Supports SD 1.5, SDXL, SD3, Flux 1, Flux 2. AI-powered model identification and prompt enhancement.
A ComfyUI pack by Maxomarai: The Last Model Switcher (one-click model switching + AI prompt enhancement) and Wayfinder (visual graph navigation aids).
Enhanced audio loading node for ComfyUI supporting .wav, .mp3, .flac, .ogg, .m4a formats with filename text output. (Description by CC)
Two simple nodes: 1. Get the steps based on the model name, 2. Generate prompts using chatglm.
GIMP color mod nodes in comfyUI
Enhanced image save node with PNG/JPEG/WebP/TIFF/BMP support, quality control, GPS EXIF metadata, and date macro expansion in filename prefix.
Professional astrology engine for ComfyUI with natal chart, transit, synastry calculation and zodiac wheel visualization nodes. (Description by CC)
Custom node for ComfyUI that integrates Puter.js AI image generation directly into workflows with text-to-image and image-to-image support.
A toon composition tool for ComfyUI. (Description by CC)
ComfyUI node for downloading videos and audio using yt-dlp with automatic quality selection and cookie support. (Description by CC)
Dataset-oriented image load and save nodes for ComfyUI with filename metadata and auto-incrementing support.
Node similar to the original videocombine with additional features: control sound volume, video timeline, and save current frame to image.
Formats datetime values in ComfyUI workflows. (Description by CC)
A custom node for ComfyUI that randomly selects a checkpoint from a configurable list of models with weighted selection, per-model CFG ranges, and enable/disable toggles.
Nodes:Masks to Colored Masks, Color Picker
Nodes:ColorListMaskToImage, FlattenAndCombineMaskImages
ComfyUI nodes enabling direct API key usage with custom proxy support for OpenAI, Claude, Gemini, and video/image generation services. (Description by CC)
ComfyUI nodes for local Whisper and Gemini/OpenRouter audio transcription for voice cloning labels.
Custom nodes for saving images/videos without metadata and GPU-accelerated frame rate processing (25fps→16fps). Includes 5 nodes: Save Image/Video (No Metadata), Frame Rate Resampler (CPU/GPU). Supports CUDA acceleration and multiple nterpolation methods. Requires ffmpeg.
Custom nodes for ComfyUI to convert between equirectangular (360°) images and a strip of 6 cube faces (90° FOV each).
Do you want to easily swap width and heigth? Than this is for you. From portrait to Landscape. Is you images model trained on 1MP pictures, then you can easily scale those down. For best pictures width and heigth should be a factor of 32, say no more...
NODES: AutoCrop, ColorQuantizer, ExtractBlack, ExtractBlackAdvanced, RasterToUniformSVG, ... (10 total)
Generates perfectly tileable/repeating patterns by circular-padding the latent tensor at every denoising step.
Adds hotkeys (F8 / Ctrl+K) and a toolbar button to cycle node link styles (Spline / Linear / Straight).
Standalone export of custom nodes for FLUX and SD3 RALU sampling experiments with model-family adapters. (Description by CC)
save your image with customized naming rule
Nodes for JSON Key/Value (KV) storage with on-the-fly UI preview, dynamic drop-downs, random selection, and inline edit mode.
Hair Edge Color Overlay node for ComfyUI that applies a soft airbrush-style color overlay to hair areas near the face. (Description by CC)
Webtoon-style pink airbrush overlay on skin areas. Simulates overlay-layer pink airbrush post-processing.
More test nodes for ComfyUI_Mira extension. (Description by CC)
Reproduces Qwen TeleStyle's style-transfer pipeline as native ComfyUI nodes with weights read directly from ComfyUI's model folders.
A collection of custom ComfyUI nodes by Mister-Link, including color parsing, frame calculation, and utility nodes for enhanced workflow capabilities.
Mithril-Nodes is a collection of custom nodes for ComfyUI that enhance workflow modularity, data routing, and configuration management. These nodes help you build more dynamic, organized, and reusable pipelines for generative AI workflows.
Apply additional sampling to specific area
It has the ability to concatenate text.
This is the node that performs the magnification calculation.
some tools.
Node: Color Enhance, Color Blend. This is the same algorithm GIMP/GEGL uses for color enhancement. The gist of this implementation is that it converts the color space to CIELCh(ab) and normalizes the chroma (or [colorfulness](https://en.wikipedia.org/wiki/Colorfulness)] component. Original source can be found in the link below.
This extension fixes URL encoding issues when ComfyUI is accessed through reverse proxies like 'jupyter-server-proxy'.
Chandra-OCR-2 is a state-of-the-art OCR model that outputs structured markdown, HTML, or JSON while preserving precise layout information from images and PDFs across 90+ languages.