ComfyUI Custom Extensions
Browse 9,420 ComfyUI extensions and 73,215 custom nodes.
ComfyUI extension with several convenience nodes.
ComfyUI extension for uploading to cloud object storage. (Description by CC)
ComfyUI workflow utilities: H3 prompt editor, media loading, wiring, switches, memory cleanup, and performance monitoring.
comfyui node for aspect ratio.
A DAELab-maintained custom node library for ComfyUI providing boolean hierarchy management, SAM3 prompt rerouting, polygon masking, and Seedream workflow support. (Description by CC)
Tile-safe helper nodes for making seamless PBR materials in ComfyUI: de-light, height from depth + albedo, normal (GL/DX), AO, roughness, metallic, ORM packing and engine-ready saving.
DeepFuze is a state-of-the-art deep learning tool that seamlessly integrates with ComfyUI to revolutionize facial transformations, lipsyncing, video generation, voice cloning, face swapping, and lipsync translation. Leveraging advanced algorithms, DeepFuze enables users to combine audio and video with unparalleled realism, ensuring perfectly synchronized facial movements. This innovative solution is ideal for content creators, animators, developers, and anyone seeking to elevate their video editing projects with sophisticated AI-driven features.
Zip selected ComfyUI output folders and upload them to a Hugging Face repo automatically when the whole queue finishes. Uploads are verified.
Terminate or stop your RunPod pod automatically after Comfy-HF-Uploader has uploaded and verified your ComfyUI outputs. Dry run by default.
Directional mask expansion node that extends masks along specified directions with pixel-precise control, plus separate pinch/widen morphology operations.
Send a ComfyUI prompt to a DGX Spark or any OpenAI-compatible model server and get text back.
A custom ComfyUI node that uploads images directly to Google Drive. Designed for serverless environments like ComfyDeploy.
Custom node to load custom anima lora
이 노드는 여러분이 한국어로 상상하는 모든 것을 멋진 이미지로 만들어 줄 수 있도록 도와주는 도구입니다. 코딩이나 복잡한 설정 없이도, 마치 전문가처럼 상세하고 멋진 영어 프롬프트를 자동으로 만들어줍니다.
Production nodes for ComfyUI: prompt sheets from CSV, a model-free quality gate, .cube LUT grading and provenance sidecars.
Upload your ComfyUI generated images and workflow metadata to Lumin.
Quality-of-life nodes for ComfyUI, standard library only. Model Name Extractor names your saved files and folders after the model that made them, for use with any save node. Save Image (Macro) does the same while saving, and adds what only a save node can: a workflow .json numbered to match each image, a per-node switch to strip the workflow out of the PNG so you can post pictures without leaking prompts, an optional plain-text summary of model, LoRAs and sampler settings, and path shortening that only acts when a path would actually exceed the Windows limit. Prompt Prefix keeps system prompts as .txt files and offers them as a dropdown, so a different prefix per model is one click instead of a paste. Concat Text joins up to ten inputs, accepts any type so LLM metadata sockets connect without a converter node, and takes a delimiter that can actually hold newlines.
Mixed FP8 and NVFP4 quantization for SDXL, Wan 2.2, Qwen Image, Z-Image, FLUX, and FLUX.2 diffusion models in ComfyUI
Native-compatible 6DoF camera paths, chaining, visualization, and multi-frame reference conditioning for ComfyUI Wan video workflows
A collection of custom nodes for ComfyUI that provide advanced face callout, annotation, and compositing effects using OpenCV and PIL. These nodes are designed for image processing workflows that require face detection, annotation, and creative compositing.
ComfyUI nodes for NVIDIA AI Vision and Text models
A bounded multimodal OpenRouter node for ComfyUI with text-only output.
Bridge to a Photopea instance for better inpainting
A collection of useful nodes for ComfyUI that provide various workflow enhancements.
Standalone fork of the ImageBlendAdvance node from ComfyUI_LayerStyle with all external dependencies removed for independence. (Description by CC)
A ComfyUI custom node that joins multiple text inputs with automatic input expansion as more nodes are connected.
A collection of essential utility nodes for handling image sequences, video, audio, and synchronized data within ComfyUI workflows.
Node to use NBP with custom API key to improve data privacy control.
Node to use NBP with custom API key to improve data privacy control.
A node for Veo3 generation for internal Uncut use
Nodes to clean up your workflow.
Custom ComfyUI nodes by Thomas Smith.
ComfyUI custom node for collecting per-layer activation statistics (Hessians and optional amax) for external quantization tools like GPTQ, OBQ, and ConvRot.
Efficiency tools, Personalized style, Other Nodes, ...
Prompt Shelf is a small ComfyUI custom node pack for reusable prompt text, named prompt fragments, resolution presets, and simple string combining.
Bundled custom node pack for the SynthID-Bypass v2 workflow.
Aimacor Labs — Model Downloader: descargador integrado de modelos, LoRAs, VAEs y checkpoints para ComfyUI, con soporte para Civitai y HuggingFace.
Aimacor Labs — Video Prompt Director: generador de prompts de video para ComfyUI, con selectores por categoría (cabeza, expresión, cámara, etc.) respaldados por librerías de frases editables.
Provides a high-precision RGB/HSV/Alpha color picker as a custom node for ComfyUI.
Gemini image generation and prompt utility nodes for ComfyUI.
ComfyUI node that opens generated images in a fullscreen inspector overlay.
Small utility nodes for ComfyUI workflows.
Ask Waxy AI Realism Test whether an AI-generated face looks real versus AI, right inside ComfyUI.
Share your local ComfyUI via a Cloudflare Quick Tunnel with an optional password gate
A minimal node set in comfyui to render pydocmaker reports
ComfyUI custom nodes for ToriiGate-0.5 with local GGUF direct connection mode. Preserves original Transformers nodes while adding GGUF model support for image description, dataset caption generation, and prompt-assisted generation in ComfyUI workflows. (Description by CC)
Advanced guidance for dual-network diffusion models (e.g. Ideogram 4) combining channelwise CFG normalization, APG, and CFG momentum.
Discontinued ComfyUI node package providing ApplyNativeFlux2SpotEdit for native Flux.2 SpotEdit-style image editing. (Description by CC)
Inpainting/outpainting, tiling and ref editing logics. Also an unofficial extension of Impact Pack.
StableArt GPU worker as a ComfyUI custom node: connects to the StableArt node broker and runs image-generation and ai-caption jobs on the local ComfyUI.
Nano Banana, GPT Image, Seedream, Seedance, Kling, Wan and chat models in ComfyUI through one APIKO API key.
Offline interactive 3D scene editor for ComfyUI
Converts Gaussian Splat point cloud data (PLY_DATA) to 3D meshes (TRIMESH) for export in GLB/OBJ/PLY/STL formats. (Description by CC)
System monitor for the ComfyUI menu bar: CPU, RAM, swap, AMD GPU load, VRAM and temperatures, read from sysfs without NVML or rocm-smi.
An implementation to detect underage subjects in images for ComfyUI.
Based on a YOLO model, it performs object detection, capture, and smooth tracking-based cropping on images.
Custom ComfyUI nodes that automatically populate dropdowns based on your folder structure and user configuration, making it easy to generate consistent file names.
This custom node provides two different approaches to automatically match text prompts with LoRA models using their descriptions.
Massive scalable SDXL/Flux prompt system with 300+ atomic options.
Create A Meme
TKVideoZoom - various video effects (zoom,slide,spin,etc), TKVideoStitcher, TKVideoFuse, TKVideoSmoothLooper, TKVideoSpeedZones.
A ComfyUI text node that samples tags from a probabilistic model with base scores and influence matrix.
Simulates authentic instant film (Polaroid/instax style) frames, physical marker handwriting captions, and adjustable chemical film emulation (0-100) in ComfyUI.
ComfyUI node pack for FireRed-Image-Edit-1.0: instruction-based image editing, multi-image composition, and text-to-image generation.
Records how many times each model (checkpoint / diffusion model / LoRA) is used, incremented on every generation.
Human-in-the-loop image selection tool for ComfyUI workflows using a Flask web server, enabling users to pause workflows and interactively select images via a web browser interface.
Apply color **LUTs** and **swatches/palettes** to images in ComfyUI from one self-contained collection.
Nodes for turning a flat photo into a VR180 image with the flat2vr (MiniMax-H3-Image) and l2r (Qwen-Image 2.1) LoRAs.
Enterprise creative validation, attention heatmaps, and campaign impact scoring by Trusapien.
Unified ComfyUI nodes for external API integration (Replicate, Fal.ai, HuggingFace)
Custom nodes for building reusable text, image, and prompt-template libraries directly inside ComfyUI.
Write prompts from a markdown rulebook: an OpenAI-compatible LLM rewrites your idea into a structured positive prompt, plus a deterministic Layout compiler. Endpoints live in skillprompt.ini.
A very practical text node. Break the prompt into words or small sentences according to punctuation marks. You can freely turn on or off each small segment. Edit, lock, insert content, adjust order, increase or decrease weights, and implement reactions in the node edit box. There are two modes: intercept mode and accumulation mode. Additionally, there are many practical tools including text format purification, text box mixer, camera prompt tool, load external text, and LLM prompt optimization. 一个非常实用的文本节点。将提示词按标点炸开成单词或小句,你可以对每小段进行自由开关、编辑、锁定、插入内容、调整顺序、增减权重,并实施反应在节点编辑框。针对上有来词有拦截和蓄积两种模式。另有格式净化、文本框混合器、相机提示词工具、加载外部文本、LLM提示词优化等多种实用工具。
A ComfyUI custom node that randomly selects one input from multiple inputs of the same type.
A ComfyUI custom node that concatenates 2-32 string inputs into one with optional delimiters. (Description by CC)
Provides custom nodes with a cooldown feature for ComfyUI. (Description by CC)
Custom nodes for ComfyUI to expand images with noise for outpainting or inpainting purposes. This allows for seamless image extension in any direction.
A librarian tool for your models, workflows, datasets and basically anything else you use.
Resolution suggestion and downscale helper node for ComfyUI.
a comfyui plugin for image to avatar
a comfyui plugin for image to comic
image to sticker
Custom ComfyUI nodes for enhanced text and image processing
Run Nunchaku 4-bit SVDQuant models on AMD Radeon and Instinct GPUs (ROCm): up to 3.6x faster than stock ComfyUI on a 16 GB RX 7800 XT. Supported models: FLUX.2 [klein] 4B and 9B, FLUX.1-schnell, dev, Krea and Kontext, Qwen-Image, Qwen-Image Lightning and Qwen-Image-Edit, Z-Image-Turbo, SDXL and SDXL-Turbo. Includes a LoRA loader for all of them, including speed LoRAs (LCM, Qwen-Image Lightning). Setup: install here and restart ComfyUI (Manager adds nunchaku-rocm, plus Triton on Windows, without touching your PyTorch). On Windows its install.py also adds the ROCm SDK headers and Python files Triton needs. Then open Templates > Extensions > comfyui-nunchaku-rocm: each example workflow lists its model downloads. The first image of each model compiles kernels for a minute or two.
A ComfyUI custom node that applies subtle, handheld-style camera shake to an image batch. It adds per-frame affine transforms (translate, rotate, scale) to simulate natural micro-movement, ideal for adding life to video/animation frame sequences.
A collection of highly functional custom nodes for ComfyUI designed to manage line-by-line text reading, progress visualization, and dynamic input switching without complex wiring. (Description by CC)
Professional broadcast and social platform resolution delivery suite for ComfyUI video workflows.
Export ComfyUI videos as DNxHR or ProRes .mov files that DaVinci Resolve (including the free version) can import, with audio and correct color.
VAE Decode node that solves VRAM problems with mathematically precise temporal stitching, maintaining perfect frame counts across batch sizes for video generation.
Advanced VAE decoder node with audio sync precision and crash-proof memory management for long videos. (Description by CC)
Interactive crop node for ComfyUI (inline preview with draggable selection).
Minimal OpenAI API nodes for ComfyUI.
UIIIAIII Toolkit: a node category manager (rename / move / hide / sort / restore categories directly in the node library tree via right-click menu and drag), node and plugin UI translation for multiple languages (derived from AIGODLIKE-ComfyUI-Translation and ComfyUI-DD-Translation with modifications), plus bundled nodes: Agnes AI (text-to-image / image-to-image / text-to-video / image-to-video / keyframe animation), Qwen-Image-2.1 (unified generation and editing) and Qwen-Image-Edit-2511 via ModelScope API, text input/preview, background fill, and random noise seed.
A ComfyUI custom node for face comparison. This node utilizes Face++'s facial recognition and comparison algorithms by directly calling the Face++ API. Its usage in the workflow is as follows:
A simple node to fetch a file and put it in a specific path
Automatic workflow synchronizer/loader for UmeAiRT Workflows (Flux, SDXL, WAN, etc.).
A custom ComfyUI node that pads an image to a multiple of 8 width.
Nodes for swapping a person or an object in a video: long motion transfer, face pass, object fit and paste-back, low-memory clip loading and finish.
Unofficial ComfyUI custom nodes wrapping proper-pixel-art "Fixes AI pixel art images, video, or sprite web uploads".
This custom node uses OpenAI Image API to generate image (if no input image is provided) or edit image (if input image is provided) with the latest gpt-image-1 model. To use it, you will need to provide your OpenAI API key. This makes the node to be friendly for situations where ComfyUi serves as API server, because you don't have to login like the official OpenAI GPT Image 1 node does.