ComfyUI Custom Extensions
Browse 9,446 ComfyUI extensions and 73,215 custom nodes.
Character prompts from reusable JSON presets: chained section nodes (hair, outfit, scene...) with per-field overrides, preset references, LoRAs and templates, plus Tansu, a preset organizer and editor tab.
Custom ComfyUI node collection providing image tools (Load Image, Color Match, Canvas Editor) and LLM nodes for integration. (Description by CC)
Custom ComfyUI node for generating or improving prompts via DeepSeek API.
ComfyUI custom nodes for RecGen 2D RGB-D to 3D reconstruction
WorkerKeeper is a background service for ComfyUI that automatically kills idle comfy-env (GeometryPack, GaussianPack, SAM3, MoGe2, Sharp, PanoPack, DepthAnythingV3, etc.) isolation subprocesses. It requires zero workflow changes — install and forget. An optional manual override node is also provided.
Path utility for ComfyUI
Text file utility for ComfyUI
A simple set of nodes to make things easier. String nodes, files nodes and random prompt nodes.
A simple set of nodes to make things easier. String Nodes and Files nodes.
Convert EPUB's into audiobooks using TTS nodes
ComfyUI nodes for HeartMuLa - Music generation and lyrics transcription pipeline
Collection of ComfyUI custom nodes for image manipulation (Expand, Scale, Color Picker)
ComfyUI plugin for reading model file metadata including download addresses and model information. (Description by CC)
A smart ComfyUI extension that simplifies text-driven workflows. Systematically organize, aggregate, and switch prompts or character settings with a single click, and batch process multiple staging plans simultaneously.
Speaker-tagged dialogue TTS nodes for ComfyUI. Currently provides an IrodoriTTS-v2 backend for generating two-character dialogue audio from dialogue_segments_json.
A lightweight, standalone ComfyUI node for LLM-powered text and prompt generation that connects to any OpenAI-compatible local LLM server.
A standalone ComfyUI custom node that sends workflows to a remote ComfyUI instance and returns the output image — all in a single node.
A standalone ComfyUI custom node that sends workflows to a remote ComfyUI instance and returns the output image — all in a single node.
A unified ComfyUI node pack: a LoRA loadout/loader, a model library, prompt management, media loader/output nodes, Config Switcher nodes and a word bank. Optional nodes live in KRC NodePack Extras.
Optional KRC nodes built on KRC NodePack: integrations with other custom-node packs (e.g. MiniMax H3 RefMods) and standalone utilities (e.g. the KRC Frame Router). Requires KRC NodePack.
Custom nodes intended to improve usability.
Utility nodes for interacting with images and image sizes
A ground-up ComfyUI inference implementation for NormalCrafter. It uses the released NormalCrafter UNet/VAE weights and the Stable Video Diffusion image encoder/scheduler.
Production-focused ComfyUI nodes, API adaptions, runtime patches and workflow utilities.
在任意多行文本框左下角挂按钮:本节点历史提示词 + 按工作流分类的预设库
Deterministic speech bubbles and readable dialogue for finished 2x2 comic pages in ComfyUI.
native batch ONNX tagging and in-process SDXL LoRA training; no WD14 node pack or TrainTools-MZ required for new workflows
A minimal ComfyUI custom node for extracting pixel bounding-box coordinates from Ultralytics YOLO detections.
Thumbnail-based visual prompt part picker and library manager for ComfyUI.
Some custom nodes for weird niche needs when working with Krita
ComfyUI nodes for batch loading images from ZIP archives, filtering by index, and saving with original filenames. (Description by CC)
Custom ComfyUI nodes for CHARON auto-alignment and DA3 camera export.
Ordered multiline prompt-list conditioning node for ComfyUI
Reusable UNET, text encoder, VAE, and multi-LoRA presets for ComfyUI
Drop-in Save Image, Save Audio and Save Video nodes that mark AI output as it's saved: an invisible watermark and signed, time-stamped C2PA Content Credentials, the AI labels the EU AI Act (Art. 50) and California ask for. Images are marked on your machine; audio and video in Kennmarq.com's cloud. Example workflows and a local safety check included.
A LoRA loader that fetches and displays Civitai metadata including trigger words and preview images, with local caching and NSFW blur support.
A ComfyUI custom node for looping through images in a folder with a visual library UI and playback controls.
Nodes: Live Chat Stream. Chat with an LLM / VLM (Ollama) inside a node and turn the streaming reply into images - generation starts as soon as the prompt is written, while the rest of the reply is still streaming. Two prompt lanes (clean <prompt> block and the model's own thinking / reply / inner monologue), character presets, image input/output for I2I, optional decision-model image judge, VRAM auto-unload, and multi-language UI (en/ja/zh).
Real-time particle animation node for ComfyUI. Renders smoke, sparks, light rays, and star-warp effects using PixiJS v7, with 13 post-effect filters, blend modes, custom textures, character-shape particles, and scatter mode. Captures any frame as an image output.
A ComfyUI custom node for looping through text prompts one at a time, with two input modes: typing directly on the node, or selecting .txt files from a built-in 3-pane library (folder tree, file list, and in-place preview/editor). Supports presets, natural/random sorting, and registering read-only external paths to reuse existing prompt/wildcard .txt data from other node packs.
Full-browser 3D scene / pose / video studio for VRM models running on ComfyUI (/vps). Multiple VRM models, props, lights, cameras and cuts on a keyframe timeline, image/video to pose with the native SAM 3D Body nodes, regional prompts (BV Node Pack), and a Generate tab that sends renders to your workflows via the Load Studio Render node. three.js / three-vrm are bundled locally (no CDN).
Drop a ComfyUI PNG/WebP/JSON onto the node to instantly extract and apply checkpoint, VAE, and prompts from embedded metadata. Supports ComfyUI, SD WebUI, SD Forge neo, Fooocus, SDXLPromptStyler, and Lora Loader (LoraManager). Includes LoRA from Metadata and CLIP Text Encode edit+ nodes.
ComfyUI node to merge a batch of masks into a single mask via max or sum_clamp
ComfyUI node to save images while copying EXIF metadata from a source image
A ComfyUI node that provides advanced image adjustment filters and controls for image manipulation
Independent multimodal prompt director with optional llama.cpp, Ollama and OpenAI-compatible providers.
Official OpenAI GPT-Image-2 API nodes for ComfyUI: Text-to-Image & Image Editing with multi-image support
ComfyUI custom node for Google Gemini Flash Image generation and editing
ComfyUI custom nodes for automated audio quality evaluation and correction (LUFS, True Peak, Noise Floor, LRA, Dropout)
ComfyUI node that sends a prompt to an LLM (Claude, OpenAI, Kimi, DeepSeek, or a local OpenAI-compatible server) and returns an enhanced version, for use ahead of image/video generation nodes.
Comprehensive single-node solution for ComfyUI integrating base generation, latent refinement, neural upscaling, and post-processing.
A ComfyUI custom node that converts video frames to side-by-side (SBS) stereo format for VR viewing.
API-first fashion generation and evaluation nodes for ComfyUI.
A suite of experimental custom nodes that allows access to Google's 1P models like Veo, Nano Banana, Gemini, Virtual-try-on, Lyria
Save Your Image to BunnyStorage
Uploads image(s) to a mood.site board as PNG, replicating the mood.site upload Apple Shortcut.
ComfyUI custom node pack for building karaoke pipelines — yt-dlp source, stem separation, and audio plumbing nodes.
A ComfyUI custom node for compressing images via TinyPNG API.
This extension offers wildcard prompting works solely in workflow.
Fetches CivitAI trainedWords (trigger tags) for a LoRA stack and returns them as a comma-separated string.
A monorepo for personal ComfyUI tweaks, frontend extensions, custom nodes, and future utilities.
Kongshan product-image workflow nodes for ComfyUI, including Gemini, CLI, SAM, GroundingDINO, file IO, and local image utilities.
Custom set of nodes used by glif.app. With glif you can build mini apps that are powered by custom comfy workflows.
ComfyUI custom node for selecting game characters with real-time preview and tag output generation. (Description by CC)
ComfyUI custom node that converts an input image into a Stable Diffusion-ready prompt using Abacus.AI’s OpenAI-compatible multimodal API (model dropdown fetched from /v1/models; supports route-llm and other available models). Requires an Abacus.AI API key and consumes API credits.
ComfyUI nodes for FLUX and SDXL image generation using Super Adapter models with HDR VAE support.
Modular collection of utilities and custom nodes designed to enhance ComfyUI, including tools for workflow automation and pipeline creation.
ComfyUI custom node for image generation with an OpenAI-compatible primary endpoint and Tencent Cloud VOD AIGC fallback on rate-limit.
ComfyUI node for building rich, conditional prompts with a variable system, weighted random selection, tag-based filtering, and Jinja2 templating.
Lightweight custom nodes that integrate ComfyUI with ForgeUI-style samplers and provide convenient ControlNet and LoRA stack helpers.
ComfyUI node for controlling local Forge UI / Automatic1111 WebUI via API, designed for InstantID + Canny ControlNet workflows on SDXL.
Simple node for ComfyUI that applies a text-to-image (T2I) style to an image or embedding with support for start/end frame control and a weight parameter.
A simple ComfyUI node that creates a white canvas for freehand drawing and automatically runs inference for sketch-to-image workflows.
A small ComfyUI custom node that cleans up prompt text and appends style terms.
Mask Canvas Editor is a ComfyUI custom node that provides an interactive visual editor for positioning a background image behind a mask region.
Standalone ComfyUI custom node for evaluating two conditions against one input value.
ComfyUI T2I 测试插件:全能加载器、手工提示词单项测试、修手重绘
Custom node for replacing unstable GPT image proxy nodes, automatically parsing proxy service returns and reporting errors instead of outputting blank placeholders. (Description by CC)
NODES: External Boolean (ComfyUI Deploy), External Checkpoint (ComfyUI Deploy), External Image (ComfyUI Deploy), External Video (ComfyUI Deploy x VHS), OpenArt Text, Image Websocket Output (ComfyDeploy), ...
Kumiho ComfyUI custom nodes - Connect ComfyUI workflows with Kumiho Cloud asset management system
Deterministic natural-language-to-prompt compiler for ComfyUI (Illustrious).
Nodes:A_Face3DSwapper, A_FaceCrop, A_FacePaste, A_OpenPosePreprocessor, A_EmptyLatentImageLongside, A_GetImageSize, AlkaidLoader, AdapterFaceLoader, AdapterStyleLoader, ...
ComfyUI custom_node that load and save file directly from S3 Simplified version of [a/https://github.com/kealiu/ComfyUI-S3-Tools](https://github.com/kealiu/ComfyUI-S3-Tools)
NODES: TextEncodeQwenImageEditKrsSimple, TextEncodeQwenImageEditKrsAdvanced, QwenImageEditLatentKrs, ...
A collection of small utility nodes for ComfyUI that make common prompt-building and sampler configuration tasks faster.
Browse, download, and use KurutAI LoRAs and prompts directly in ComfyUI.
Crossbreed models of the same architecture, block by block.
A ComfyUI custom node for creating chronophotography effects from video frames
GoRi ComfyUI custom nodes — Camera Director (camera-aware prompts and conditioning for Qwen-Image 2.1).
In addition to achieving conventional single-image, single-round reverse engineering, it can also achieve single-image multi-round and multi-image single-round reverse engineering. Moreover, the Phi model has a better understanding of prompts.
ComfyUI custom node for splitting blue-pencil manga manuscript pages into story panels.
OpenToonz-inspired cleanup, gap closing, and palette-region ComfyUI nodes for rough blue manga manuscripts.
A ComfyUI custom node that weights tags based on their frequency/rarity in danbooru
Pixel optimization tools for ComfyUI image processing. (Description by CC)
A node to pause the workflow to select one of the three generated images for further processing.
EARLY ALPHA: optional local inference companion for Genereti and ComfyUI. Native macOS requirements; explicit model downloads.
WAN Dual LoRA Loader
Utility, common, and experimental nodes for ComfyUI.