ComfyUI Custom Extensions
Browse 9,446 ComfyUI extensions and 73,215 custom nodes.
Save and stream intermediate sampling steps during diffusion. Perfect for showing generation progress on your frontend instead of a loading spinner.
Comprehensive prompt management system for ComfyUI with database storage, thumbnail management, advanced search, and cross-platform support.
Adds two nodes that reuse upstream ComfyUI-QwenVL presets but add a runtime override that can be wired/unwired without getting stuck.
Auto-detects the background haze in dark images and crushes it to true #000000 for OLED wallpapers, while protecting faint detail like smoke, hair and anti-aliased edges.
A custom node for ComfyUI that allows you to format strings using Python's string formatting syntax.
A lightweight ComfyUI node library providing logic operators, type conversions, math utilities, and switch nodes.
A ComfyUI custom node plugin that calls Google Gemini-compatible REST endpoints for image analysis, text-to-image generation, and image editing.
A custom ComfyUI node that provides Flux 2–recommended resolution presets for quickly creating empty latents, without manually entering width and height values.
Make your BrandMoves pictures on your own computer, for free. Connect ComfyUI to BrandMoves with one code.
Crop and Paste custom nodes for ComfyUI — crop images to mask bboxes and paste them back with alpha blending
ComfyUI custom nodes by Daply (image loader from URL, and more).
Advanced empty latent node for ComfyUI with 67 resolution presets, aspect ratio filters, manual sizing, and model-aware latent formats.
Privacy-aware image loader and input folder organizer: gallery, clipboard, move, recycle and subfolders.
Prompt editor and visual LoRA manager for ComfyUI
Browse, search, preview, and download models from Civitai and Hugging Face directly inside ComfyUI. Features include dark/light themes, keyboard navigation, NSFW blur, Prompt Fetcher node, local model management with 29 folder types, SHA256 hash verification, and comprehensive animations.
Custom ComfyUI nodes by Orel for downloading Hugging Face models, managing HF tokens, and uploading LoRAs.
A complete set of custom nodes for ComfyUI that integrates the Qwen2-VL vision-language model with all-in-one QWEN VLM MACHINE supporting Transformers and GGUF loading, plus image and video support.
ComfyUI custom nodes for preserving and extracting workflow metadata
Combine ordered STRING inputs and native lists with optional inline text preview and daily per-workflow logging and persistent input labels.
Audit character consistency across AI-generated image batches using Claude's vision API.
Prompt optimizer for FLUX.2 [dev] and Qwen-Image 2.1: calls an OpenAI-compatible LLM with a prompt-engineering spec (skill), plus a free local prompt checker.
ModelConfigSelector node for managing checkpoint configurations with manual override support and batch sampler parameters. (Description by CC)
Nodes: DSP Image Concat
Two lazy-evaluation switches: an Anima flow switch that flips model, positive and negative with a single dropdown, and a ControlNet switch. When skipped, the whole upstream branch is never executed at all.
Five switch nodes that keep the graph running when an upstream node is muted (Ctrl+M): mute-bypass, A/B, multi-any merge, multi-channel merge and output fanout. Pure reference hand-off - no copying, conversion or device moves.
Futurematrix ComfyUI Custom Node
Custom comfyui nodes for my purposes.
Highly flexible multimodal and reasoning-aware custom node suite using Hub and Spoke architecture to connect ComfyUI to any OpenAI-compatible API with advanced sampler injection.
EXR inspection and A/B QC for renders and AOVs: wipe/blink/diff, exposure, false color, eyedropper with exact float values. Multilayer EXR (Nuke/Blender) decoded on the server, every compression, overscan, data passes.
deAPI nodes for ComfyUI: low-cost cloud text-to-image, image edit, background removal, upscaling, text/image-to-video, text-to-speech, music and transcription. No local GPU needed.
A slightly cursed ComfyUI custom node that creates pruned copies of images without ComfyUI metadata.
A CLIP prompt compiler that folds negative prompts into positive prompts with NegPip negative weights for guidance-distilled models.
ComfyUI nodes for DeepGen API
A node for saving audio in MP3-format with selected bitrate mode and quality to an output subfolder or absolute path on any drive.
In-app documentation for 3500+ ComfyUI nodes, in English and Chinese. The doc panel language follows ComfyUI's Comfy.Locale setting.
Low-VRAM and 16GB-RAM optimization node for ComfyUI native Qwen-Image-2.1 (chunked SwiGLU FFN, head-chunked Attention, prefix KV cache control, and memory cleanup).
Top-bar real-time VRAM, RAM, and CPU monitor with one-click GPU memory cleanup for ComfyUI.
A lightweight, local, open-source image editor (miniPaint) seamlessly integrated into ComfyUI. Supports both iframe and new-window modes, full compatibility with Node 2.0.
Video Merge Node that combines image batches with SRT text, audio padding, and mixes voice and background music into final video output.
This node adds an outside-only outline around the opaque region of an image while cleaning up semi‑transparent edge pixels and removing stray islands.
ComfyUI custom node that converts input image dimensions to the nearest recommended Qwen-Image resolution preset, matching orientation and aspect ratio while optimizing for the target model.
This node receives an int and returns the positive version of it in any case.
Frame-accurate layered video editor inside ComfyUI for stitching AI-generated clips into one continuous piece. Auto Stitch recommends the best cut point between two clips using frame and motion matching; also scene/beat detection, version stacks, wipe compare, pre-render, and MP4/MOV/WebM export.
A simple ComfyUI node development starter template.
ComfyUI custom nodes for Hi3D Image-to-3D API: generate 3D models (GLB) from images.
Nodes to convert data such as images and audio from Base64 format into comprehensible format for use with other ComfyUI nodes.
StyleGAN2 and StyleGAN3 support for ComfyUI, @dfl's fork with added features
Context-preserving single-pass inpaint sampler for ComfyUI.
Lightweight model-agnostic compositor for preserving source regions and blending generated edits in ComfyUI.
ARC-focused Gemini / Nano Banana image generation and editing nodes for ComfyUI, derived from ComfyUI-NanoB-Edit-Gemini.
ARC-native model-only multi-LoRA loader for ComfyUI.
Gemini and OpenAI image generation and editing node for ComfyUI.
Lightweight model-agnostic compositor for preserving source regions and blending generated edits in ComfyUI.
Inline image preview, transform, and mask painting node for ComfyUI.
Clipboard-friendly image and mask loader for ComfyUI.
Advanced image masking and editing node for ComfyUI.
Timeline director and companion nodes for MiniMax H3 in ComfyUI.
Gemini / Nano Banana image generation and editing nodes for ComfyUI, derived from ComfyUI-NanoB-Edit-Gemini.
Manual Smart Memory control nodes for ComfyUI.
Keep unused ComfyUI workflow branches dormant until they are needed.
Inline image preview, transform, and mask painting node for ComfyUI.
Manual Smart Memory control nodes for ComfyUI.
Keep unused ComfyUI workflow branches dormant until they are needed.
ComfyUI custom nodes for angle-based image rotation with high-quality interpolation and lossless restoration
A custom node for ComfyUI that allows selecting a base resolution, applying a custom scaling value based on FLOAT (up to 10 decimal places), and adding an extra integer value. Outputs include both INT and FLOAT resolutions, making it perfect for you to play around with.
Custom utility nodes for ComfyUI by Digital Divas.
Custom ComfyUI nodes for Nano Banana Pro image generation and editing via fal.ai API.
A lightweight ComfyUI node for comparing two images in one preview.
ComfyUI extension that repairs moved model paths by finding exact filenames across registered model directories.
Roto / alpha-matte helper nodes for ComfyUI: save an RGBA image sequence as one TGA zip named after the source video, batch-load a folder of folders of videos, a spec readout (size / frames / fps), and a preview-only roto review player (wipe / alpha / history / download).
ComfyUI nodes for Pulse-of-Motion - a Visual Chronometer that predicts Physical FPS (PhyFPS) from video frames.
ComfyUI custom nodes integrating CosyVoice 2 and FishSpeech TTS engines with flexible model configuration and audio routing. (Description by CC)
ComfyUI custom nodes library for audio pipeline with FFmpeg batch conversion support for converting videos to audio formats. (Description by CC)
A collection of little custom nodes for ComfyUI to help in simple/specific tasks
A small collection of ComfyUI audio nodes, starting with peak and RMS normalization.
A set of PoC ComfyUI cloud API nodes for connecting the cloud API with a local ComfyUI instance
Brightwire utility nodes for ComfyUI video/keyframe cache workflows.
ComfyUI custom node for Trellis mesh normal postprocessing.
Tiny type adapter for workflows where one node outputs a hex color string and another node requires a color input.
A tiny ComfyUI custom node pack for compositing two transparent images like Photoshop's normal layer mode and saving as transparent PNG.
Smart offline checkpoint loader that auto-detects model architecture and lineage, dynamically routing optimal Latent, base prompt, and VAE warnings. (Description by CC)
A collection of custom nodes for ComfyUI that provide various input and output capabilities.
Nodes:CRE8IT Serial Prompter, CRE8IT Apply Serial Prompter, CRE8IT Image Sizer. A few simple nodes to facilitate working wiht ComfyUI Workflows
Local-only prompt receiving, queue control, and image reporting nodes for ComfyUI.
custom node which allows placing **some_tag** in your prompt and it will replace it with a preset prompt as you like.
Professional navigation bar widget for ComfyUI with JWT-based user authentication, workflow templates, and team collaboration features in a purple gradient design.
A streamlined ComfyUI custom node package for interacting with GROQ's AI models, including both text and vision capabilities.
Experimental ComfyUI custom nodes for applying prompt conditioning to color-painted mask regions by converting color mask images into ComfyUI tensors.
Load images from a pasted filesystem path in ComfyUI. The loader is a structural clone of the built-in Load Image node, so the mask editor, clipspace and inpainting all keep working; a companion node loads whole folders in place.
A node suite for prompt engineering with persistent character and segment libraries.
Filter LoRA trigger pollution + RegionalLayoutMasks (!!/!!!/TB/TMB) for Attention Couple workflows
A ComfyUI node for interfacing with an OpenAI‑compatible TTS API endpoint.
批量多API图生图节点,支持社区版/全能Xinbao/官方PRO混合策略,支持OSS图床与上传URL输出
Load image from URL for ComfyUI - supports single image download and convert to ComfyUI IMAGE tensor
This node performs detection and segmentation, as well as upsampling, sampling, and blurring the composite.
Simple yolov8 comfyui plugin. This plugin adds Yolov8DetectionSegmentation node and ImageCompositeBlurred node.
Custom nodes for FASHN Human Parser, a SegFormer-B4 model fine-tuned for human parsing with 18 semantic classes optimized for fashion and virtual try-on workflows.