ComfyUI Custom Extensions
Browse 9,420 ComfyUI extensions and 73,215 custom nodes.
Lightweight ComfyUI custom node pack that plays audio notifications at different workflow execution stages on the backend machine.
Filter LoRA trigger pollution + RegionalLayoutMasks (!!/!!!/TB/TMB) for Attention Couple workflows
Run Reanimator generations on your own GPU. Your media never leaves this machine.
Personal ComfyUI custom nodes.
A ComfyUI custom node for batch-writing up to 100 prompts with per-prompt toggles, trigger prefix, folder import (txt/csv/json) and folder-based presets. Splits one run into N queue items. Bilingual UI (Chinese / English).
A simple utility node that counts active Video S2V Extend nodes in your workflow and returns that number plus one. Now you don't have to remember to update the batch number. Just set your extend nodes and go. It was easy to forget so I made this quick and dirty.
Visual camera-angle LoRA control and smart aspect-ratio resizing for Qwen image editing in ComfyUI.
Advanced image resizing node for ComfyUI with aspect ratio preservation and mask support
Real-time natural language prompt translator using Ollama Cloud API with support for English, Japanese, Chinese, and Korean. (Description by CC)
Serverless-compatible fork of ComfyUI custom nodes connecting to Kie.ai API for image and video generation with API key passed as node input for serverless environments.
ComfyUI custom nodes: (1) Pad Image by Aspect for Outpaint — expand, pad, and mask images to fixed or randomized aspect ratios with precise spatial and scale control for outpainting and canvas expansion. (2) Image Mask Comparer — compare masked regions of two images with similarity scoring and automatic retry/re-queue on mismatch.
Professional ComfyUI node pack for AI-powered fashion visualization. Transform flat sketches into product photos, ghost mannequin views, editorial shots, e-commerce images and color variants using Gemini AI.
ValiTools node collection: VSmartPrompt (Dynamic Prompts with rich-text editor, reusable variables, switcher conditions, chaining inputs and post-run selection highlighting), VFileRandom (no-repeat random image loader) and VRandomSelector (lazy random passthrough of N same-type inputs) and VWaitForVRAM (hold execution until enough VRAM is free).
Professional temporal stabilisation engine. Eliminates flicker, luminance drift, and frame-to-frame inconsistency in AI-generated and traditional video footage.
Native ComfyUI integration of SEGA-style Qwen-Image sampling.
Nano Banana, FLUX.2 Pro and GPT Image workflows: generate, edit references, queue prompts and compare current image rates through API429.
ComfyUI nodes for GPT-Image text-to-image and image editing through the Vergora API
ComfyUI node for minimax-h3 video generation through the Vergora API
ComfyUI nodes for Wan 3.0 video generation through the Vergora API
NODES:Scale Image Node, Switch Image Node, Switch Text Node, First Number Node, Mirror Effect Node, Text To ImageNode, Flow Nodes, Simple Save Image Node
Click the ⚡ button to render your open ComfyUI workflow on a Spark Fuse cloud GPU and pull the result back into your graph. No CLI, no file export.
A custom node for ComfyUI that downloads images from Pinterest boards with real-time image preview on the node and progress tracking.
One node for ComfyUI that takes an image with alpha after background removal and returns either shadow only or outline only.
Custom SamplerCustomAdvanced node with the same and new features as KSampler.
A collection of custom ComfyUI nodes developed by Victor Alves.
VideoGen storyboard-to-video nodes with native ComfyUI video output.
Vidia Open Studio video saver custom node for ComfyUI: watermark, ffmpeg mux, R2 upload, backend notify.
Headless ComfyUI nodes using a fixed loopback cinematic-audio runner service
A single ComfyUI node that recombines two images in CIE-LAB space: the L (lightness) channel comes from one image, the a/b (colour) channels come from another.
A ComfyUI custom node that provides quick access to standard image resolutions for AI image generation.
ComfyUI custom node for xAI Imagine API.
A ComfyUI custom node to upload LoRA models to Hugging Face Hub.
Visual slider, toggle, seed, and slider bank controls for ComfyUI with a lot of customization for controlling things like the min and max values of each node.
ComfyUI custom nodes for local LLM text generation
Canvas Preset Studio provides an easy way to select, preview, and validate image and video canvas sizes for AI workflows.
Addon for ComfyUI that adds 'Fix node (recreate + keep inputs)' context menu option
Runtime patch that fixes the shape-mismatch crash when combining Flux Kontext ReferenceLatent with any ControlNet node. No core files modified. Idempotent — skips if upstream fix (PR #9180) is already present.
A template for building ComfyUI extensions using React 19, TypeScript, and Vite.
ComfyUI nodes to simplify my tiled render
A custom node for ComfyUI that sends notifications (text and images) to a Telegram chat via a bot.
Lightweight ComfyUI text nodes for reusable token dictionaries and prompt filtering.
Custom nodes for ComfyUI to improve promts and image settings
ComfyUI custom nodes for VNTANA 3D content management platform
A ComfyUI node that resolves __wildcard__ tokens in prompt strings from text files.
Ragha Preview Video: an on-node video player for ComfyUI with sound, A/B compare (slider or side by side) and passthrough outputs. Streams 4K with little RAM; nothing written to disk.
Motion Verdict: is an AI video really animated, or only a camera move over a still picture? Removes the camera's pan/tilt/zoom/roll, measures what still moves on its own, and returns a verdict (animated / partly_animated / camera_only / still), an 'animated' BOOLEAN for gating seed batches, a subject-motion score, stall times, a timeline IMAGE and a JSON report. CPU friendly, works in API runs. Optional in-memory preview on the node.
Ragha Preview Video: an on-node video player for ComfyUI with sound, A/B compare (slider or side by side) and passthrough outputs. Streams 4K video with little RAM and writes nothing to disk. Inspired by kijai's Fast Preview.
ComfyUI_subjectselector is a custom ComfyUI node that allows you to manage and select text-based subject descriptions directly from the workflow UI. This node was designed to pair seamlessly with [a/ComfyUI-textsplit](https://github.com/vsaan212/Comfy-ui-textsplit), providing a clean, modular way to feed descriptive text prompts into your generation pipeline.
A custom ComfyUI node that splits text into multiple outputs for feeding complex multi-scene renders. This node allows you to dynamically control the number of splits and use custom separators.
Upload images to a remote server with ComfyUI.
A lightweight custom node for ComfyUI that checks whether a specified URL is reachable and optionally blocks workflow execution.
Mega LoRA Loader (MLX + Torch): LoRA stack in bypass mode with search and a Civitai / LoRA Manager info dialog per row. Works with Nodes 2.0.
ComfyUI nodes for running GGUF quantized Qwen2.5-VL models using llama.cpp
Post-generation garment recoloring nodes for VTO pipelines. LAB-colorspace recolor with pattern/shading preservation, plus auto-harmony palette remapping for multi-color prints.
Browse local image folders inside ComfyUI: waterfall grid, hover to read the matching prompt file, click to select, then output images and prompts one by one. Fully offline, bilingual UI.
MSST WebUI audio source separation nodes for ComfyUI (vocal/instrumental split, multi-stem, denoise, dereverb, karaoke)
A custom node collection for ComfyUI, offering enhanced image processing features.
MiniMax H3 Image text-to-image and multi-image editing through the AMD gateway. Register for an API key; no local GPU required.
Free Qwen-Image 2.1 and 2.1 Turbo text-to-image and image editing through the AMD gateway. Register for an API key; no local GPU required.
Convert images to pixel art using ONNX models.
Animation via tick. Parameter manipulation with wave generator. Unary and Binary math support. Value convert int/float/bool, VectorN and Image, Mask types. Shape mask generator. Stack images, do channel ops, split, merge and randomize arrays and batches. Load images and video from your folders. Dynamic bus routing. Flatten, crop, transform; check colorblindness or linear interpolate values.
Custom node to exit ComfyUI. Exit ComfyUI after a specified number of seconds.
This is a collection of custom nodes required for processing batch images.
Download and load multiple ComfyUI model files from Hugging Face, Civitai, and HTTP URLs.
A standalone ComfyUI node that bridges Ideogram 4 model latents to the connected Flux.2 VAE normalization.
Prompt text nodes for ComfyUI with commented-out lines: # and // line comments and /* */ blocks render in red and are stripped from the output.
Book Cover Finder tool that wraps openlibrary.org
A collection of utility nodes for ComfyUI, including useful functions such as audio processing and text manipulation.
Use your own ComfyUI workflows as generators in Weekend Studio. Pick the settings to show; results land straight in your document.
ComfyUI nodes for running GGUF quantized Qwen2.5-VL models using llama.cpp
ComfyUI 強大的萬用提示詞編輯、組合、替換、分割、刪除功能,檔案名稱自訂管理以及繁體/簡體中文轉換工具。 powerful wildcards prompt editing, combining, replacing, splitting, and deleting functions, custom file name management, and a Traditional/Simplified Chinese conversion tool in ComfyUI.
ComfyUI custom nodes for OpenAI-compatible image generation and editing endpoints.
Provides a structured model and helper nodes for organizing output files in ComfyUI.
Local IndexTTS 2.5 voice cloning, multilingual speech synthesis, and emotion control for ComfyUI.
Upscale video & images across up to 32 GPUs in ComfyUI. Split frames, pin each branch with GPU Init → Start (gpu_id 0–31), run Upscale-with-Model in parallel on CUDA or ROCm, then merge. Built for multi-card rigs — 2, 8, 16, or 32 GPUs. ESRGAN / UltraSharp / RealESRGAN.
Node Pack mostly for manipulating strings and integers
Native Intel Arc Battlemage (B70 / XPU) acceleration for ComfyUI on Windows (ComfyUI-OmniXPU adapters and diagnostics). Comfy Registry installs require one extra setup step for native XPU binaries plus --enable-dynamic-vram --reserve-vram 4; see repository README. Benchmarks in the repository.
Live Intel Arc (XPU) performance advisor for ComfyUI. Runs deterministic checks against your active environment, loaded models, and runtime logs to flag unoptimized settings, FP8 emulation bottlenecks, Shared GPU Memory spills, OOM causes, and CUDA-only incompatibilities directly in the UI.
ComfyUI nodes for saving audio, images, and video with exact format and quality preservation. (Description by CC)
NodeRig is a full 3D pose and skeleton editor integrated directly into ComfyUI as a custom node.
One node combining camera angle, lens, depth of field, film filter, and mood controls with dropdowns for ComfyUI workflows. (Description by CC)
DetailX cloud video super-resolution with quoted billing and resumable uploads
ComfyUI custom node providing dynamic I/O gateway nodes for dashboard/studio integration.
Text-only LLM node for OpenAI, Gemini, Anthropic and OpenRouter.
Custom nodes to convert images to pen plotter files
Use an online large language model to describe images.
Load presets from json file to run a set of experiments in a workflow with different models and settings
Simple custom node for ComfyUI that escapes parentheses in strings to treat them as literal characters instead of syntax.
ComfyUI custom node that downloads Chinese/English character lists with thumbnail cards, supports click selection and search, and outputs selected character data with origin grouping. (Description by CC)
AI-HIVE 图片与视频生成节点,覆盖电商、营销、广告工作流 / Image and video generation for ecommerce, marketing and advertising.
Masked local edit node for ComfyUI. Crop a user-painted region, send it to an external image editing API, and blend it back into the original image.
Nixie-tube styled world clock and inference timer node for ComfyUI. Drag it into a workflow to show a glowing vintage vacuum-tube clock + inference timer that follows the canvas and auto-scales.
A comprehensive Black & White Fine Art architecture processing suite for ComfyUI
ComfyUI custom node collection providing text processing, regex extraction, JSON parsing, and AI chat capabilities with 22 nodes. (Description by CC)
This custom node for ComfyUI allows you to hide a video file within a PNG image using a file concatenation technique. The resulting image can be viewed as a normal picture, but a standard unzip tool can extract the hidden video from it.
Custom nodes by X4Lch3mist: Free Resolution Selector (freely configurable aspect ratio, megapixel target, width/height outputs), Load Checkpoint (Auto VAE/CLIP) with per-file and per-architecture VAE/CLIP mapping, Dither 1-bit (stripes, Bayer, halftone, error diffusion) and ASCII Art.
This node can store a route with a chosen ID then load it anywhere in the workflow. Goal it to make linking less messy in my taste.
Frontend-only ComfyUI nodes for tidy, switchable workflows: wireless Pipe To / Pipe From (also across subgraphs), a LoRA Multi-Stack for comparing LoRA combinations, and Crossroads Split / Join for choosing which path of a workflow runs.