ComfyUI Custom Extensions
Browse 9,446 ComfyUI extensions and 73,215 custom nodes.
An unofficial ComfyUI implementation of Just-in-Time (JiT) for Flux.2, a training-free spatial acceleration method for diffusion transformers.
Spectrum feature forecasting for native ComfyUI Krea 2 image models
Native ComfyUI SDXL Spectrum forecaster node with proper internal U-Net integration
A ComfyUI custom node implementing TIDE (Text-Informed Dynamic Extrapolation with Step-Aware Temperature Control) for diffusion models.
An advanced ComfyUI extension that enables multi-agent LLM conversations using Ollama models.
Detects monophonic pitches in audio clips and renders ASCII guitar tabs with multiple tuning support. (Description by CC)
Local Fish Audio S2 Pro GGUF speech generation in ComfyUI via s2.cpp, with multi-speaker and voice-management nodes.
Generate high-quality images from text prompts with excellent Chinese text rendering,Edit images using natural language instructions..
A lightweight node search and logical navigation tool for ComfyUI. (Description by CC)
X-WIDE Image / Latent Sender & Receiver —— 独立的图像与潜空间收发节点(中英双语界面,修复预览恢复、草稿配额与缺失文件报错)
X-WIDE Plugin & Model Manager —— read-only-first local manager for ComfyUI custom nodes and models: startup timing, plugin dependency analysis, workflow usage, model inventory; the only filesystem write is disabling a plugin by renaming its folder to .disabled
X-WIDE Use Everywhere - a fixed fork of cg-use-everywhere: fixes UE virtual links and the flowing-link animation not being drawn in the browser.
ComfyUI custom nodes for Qwen-Image-Edit outfit/clothing edits — toggles, presets, and identity-preserving prompt builders. Useful for fashion edits, costume changes, and outfit swaps.
A powerful frame-by-frame video mask editor for ComfyUI with painting, auto BBox, hole filling, blur/feathering, and video slicing features.
Use JILIGULU GPT-Image inside ComfyUI.
Use JILIGULU Nano Banana models inside ComfyUI.
Load Image that lists sub-directories of ComfyUI/input, with a folder dropdown, a root-only filter, working sub-directory previews, cyclic arrow navigation, and a lightweight built-in mask editor.
读取图像尺寸与比例(宽高比、百万像素可直接连原生分辨率选择器),并把任意图像对齐到模型合法的倍数尺寸。Image info & size normalize nodes for ComfyUI.
This is a quick-and-filthy wrapper of [a/min-dalle](https://github.com/kuprel/min-dalle) for ComfyUI. min-dalle downloads and loads the actual model itself, making for a very simple, but very non-idiomatic, ComfyUI node.
Flow-SDE Sampler node: controls how much noise an SDE sampler injects at each step for flow-matching models at low step counts. Research code; measured on Anima, Flux and MOSS-Image, with mostly negative results documented in the README.
Ultra Image + PiD: ComfyUI multi-model image generator with LoRA and optional PixelDiT upscale.
Qwen 2511/2.1 and MiniMax H3 Image multi-reference editor for ComfyUI, with optional Qwen PiD upscale
烟火的 ComfyUI 实用节点集:文本处理 / 批量替换、音频拼接与音频批次、图像批次、视频对比、整数控制、组忽略管理等 16 个节点,中文界面,含前端扩展。
A custom node for ComfyUI designed to simplify the 'High-Resolution Fix' process. This node takes an input image, performs a direct upscale, and then re-encodes it into a latent image, preparing it for a final sampling pass to add details.
RDNA 3.5 attention research nodes for ComfyUI, including fixed-block, exact Triton, FlexAttention, and PISA paths with model-local fallback handling.
Create a custom node to select the closest images semantically to an input prompt
A ComfyUI custom node that inserts images into videos.
Nodes: Get Danbooru, Tag Encode
Simplified all-in-one nodes for LTX Video generation with Director workflow.
This is a custom node designed for ComfyUI to create transition effects between two images and generate a sequence of video frames.
ComfyUI custom node - quick empty-Latent generator with size presets (bilingual CN/EN). Outputs Latent + width/height.
Candidate racing for ComfyUI-H3-Motion-Context on ComfyUI-Distributed: workers generate alternative takes of the next clip in parallel, Master saves each with its latent embedded, and the chosen take continues the chain.
This tool detects the models required by your workflow in the ComfyUI interface and provides similar download results from civitai.com/models and huggingface.co, or Google search results (Manual open new window tab)
ComfyUI custom nodes for Kiwi-Edit video editing.
ComfyUI custom nodes for local LingBot Video inference.
A ComfyAI submit node to upload images to DoubTech.ai
Nodes: Submit Image (Parameters), Submit Image. A collection of loaders that use a shared common online data source rather than relying on the files to be present locally.
Saves any input image as JPEG to the ComfyUI output directory (selectable .jpg/.jpeg extension, quality, background fill for transparent pixels, timestamp filename prefix).
You can save and load queue tasks in json file format.
Run takecheck's acceptance checks on a generated take: frozen frames, duplicate frames, blank frames, flicker, motion, subtitle residue and audio. CPU only.
ComfyUI nodes for Kimi-VL model.
Make Multiverse avialbe in ComfyUI. Multiverse: The First AI Multiplayer World Model. Two human players driving cars in Multiverse.
Batch LoRA comparison sampler for Anima/Cosmos models
A single ComfyUI node that sends an image + text prompt to an OpenAI-compatible vision LLM and returns text. API keys stay server-side.
Named wireless results and an unlimited lazy stage orchestrator for ComfyUI
ComfyUI custom nodes for image loading with sub-directory support and original filename tracking.
Enhanced Load Image node with filename output, filename tracking, and subfolder listing
Pyscenedetect nodes for ComfyUI
AI_Text_Comfyui is a custom node for ComfyUI that connects to the OpenAI Chat API and automatically generates creative text prompts for AI workflows. This simplified version removes external dependencies like dotenv, requiring the OpenAI key to be set using a system environment variable.
Browse, download, and manage AI models, LoRAs, and workflows from CivitAI and HuggingFace, with custom nodes that carry their documented trigger words.
Output Fake Depth Maps with Detailed Object Boundaries
Collection of personal nodes including tavern card-related nodes, prompt manipulation related utilities, and a node to combine a list of strings back into one string.
Simple and lightweight type conversion nodes for ComfyUI.
ZZNodes tools for ComfyUI: Windows shutdown after 30 seconds of queue idle, with cancellation and on-node logs.
Automatically disables Load Image nodes without images, useful for workflows with optional reference images. (Description by CC)
Custom ComfyUI node for calling ToAPIs image generation interface with support for text-to-image and image-to-image operations. (Description by CC)
A custom node for ComfyUI to handle video frame sequences.
Tutu-maintained Krea2 Turbo prompt-adherence enhancer for ComfyUI
nodes for nightly moondream2 VLM inference supports only captioning and visual queries at the moment
Comfyui-SceneDetect is a ComfyUI custom node that uses PySceneDetect to locate scene boundaries in a video and emit one representative frame per scene as an `IMAGE` batch. It also returns per-scene metadata as JSON (`STRING`) and the total number of detected scenes (`INT`).
Two ComfyUI custom nodes providing image encryption and decryption with password-derived SHA-256 keying, block permutation, and pixel diffusion.
ComfyUI node: realistic, adaptive film simulation for photographic and cinematic looks.
智绘灵箱 - 功能丰富的 ComfyUI 自定义节点插件,提供图像处理、画板绘制、文本操作等多种实用功能
Starter ComfyUI custom node pack for Dustin's reusable utility nodes.
High-quality image and video upscaling using Apple MetalFX. Requires macOS 13+ and an Apple device with Metal support.
ComfyUI-ASV-Nodes make prompting easier.
A ComfyUI custom node for generating images using the OpenRouter API. Create high-quality images from text prompts and reference images.
ComfyUI Zura Studio for character performance with LTX 2.5, MiniMax H3 and Wan, H3 head replacement and viewport-to-final workflows.
A simple ComfyUI custom node that lets you apply multiple LoRAs (up to 10) to a model in one node.
Small, high-compatibility helper nodes for ComfyUI input validation, resolution selection, size presets, and basic workflow settings.
Custom text tools for helping with multiple prompt generation in ComfyUI. These tools allow more variety than just relying on the randomness of the image generator. You can create related, themed prompts, random each time.
A lightweight annotation overlay for ComfyUI.
Comfy NEKONOTE extensions
Pixel Socket Extensions for ComfyUI
Custom node system with Windows companion app for consolidating multiple text attribute dropdowns into single comma-separated output blocks in ComfyUI workflows. (Description by CC)
A custom node for ComfyUI that provides video loading and image batch splitting functionality.
配音出厂质检闸 AudioQC:逐句语速/杂声(no_speech)/吞字对账/太平提示,PASS-FAIL-WARN 判决+JSON 报告。Pass-by-door QC gate for TTS/dubbing audio in ComfyUI.
An elegant, robust, and highly-optimized ComfyUI custom node package for the tiiuae/Falcon-Perception vision-language model for open-vocabulary grounding and instance segmentation.
ComfyUI node for integrating with HP-Gemini-Browser-API to send content to Gemini web version and retrieve text and image outputs. (Description by CC)
Wraps around audio-separator command-line tool to generate stems with full control of parameters and ports.
Supports 2D dropdown selector for characters and actions that automatically combines prompt descriptions. (Description by CC)
ASCII art generator for ComfyUI with multi-language character set support
A powerful and enhanced ASCII art node for ComfyUI, based on FL_Ascii from ComfyUI_Fill-Nodes but completely rewritten with improved features and better usability.
ComfyUI custom nodes for inline variable wildcards. (Description by CC)
A focused JPEG-saving node for ComfyUI with 4:4:4 chroma subsampling and workflow preservation
ComfyUI nodes for face age detection and age-conditioned prompt building
ComfyUI node that draws bounding boxes + English word labels for vocabulary learning cards
ComfyUI custom nodes for Tencent's Hy-MT2 translation model family (1.8B / 7B / 30B-A3B).
ComfyUI nodes for Microsoft Lens / Lens-Turbo (3.8B T2I)
ComfyUI custom nodes for Xiaomi MiMo-V2.5-ASR speech recognition
Simplest Media Manager and other helping nodes for ComfyUI: browse, preview, trim and crop images, video and audio, then output them as lists.
A suite of ComfyUI custom nodes built around Google's Gemini API for image generation, spatial understanding (object detection & segmentation), and fast local utility nodes for images and masks.
A ComfyUI custom node for photorealistic virtual try-on generation using Google's Gemini API with flexible image inputs and customizable prompts. (Description by CC)
ComfyUI-EsesImageEffectBloom provides a powerful bloom image post-processing effect for Comfy. This tool leverages GPU capabilities to deliver optimized blur effect calculations, enhancing your images with a stunning glow. Whether you’re a developer or a digital artist, this effect can elevate your visuals effortlessly.
Use gemini to describe images, enhance prompts, etc.
A powerful ComfyUI custom node for optimizing and compressing images with support for multiple formats (JPEG, WebP, PNG) and configurable quality settings.
This is a Comfyui node bot for controlling comfyui via telegram bot
Human-in-the-loop bridge between the ComfyUI canvas and local AI agents: consent-gated graph proposals, labelled N-way switch, structured feedback, report forwarding and opt-in canvas sharing.
Dynamic Ace Data Cloud chat model IDs and completions for ComfyUI
Face Transform Dify-parity tools through Ace Data Cloud for ComfyUI