ComfyUI Custom Extensions
Browse 9,446 ComfyUI extensions and 73,215 custom nodes.
Custom Character Voice nodes for ComfyUI.
Glass-aware 16-bit RGBA hero images for bottles: cylindrical label wrap and lock, plate alignment, glass band masking and difference matting.
only show dropdown of loras ina a given subdirectory
Custom node for sending a ComfyUI image to ChatProvider Google AI/Gemini and returning the model response as text.
MOSS multi-precision inference, VRAM budgets, weight-transfer scheduling and ComfyUI workflows. Uses KohakuMoss inference components.
A ComfyUI extension that tracks random seeds throughout your image generation workflows
Productivity custom nodes for ComfyUI: image list scheduler, auto image list loader, Windows folder picker, multi-garment VTON controller, canvas tag muter/bypasser, and ready-to-use VTON example workflows.
ComfyUI custom node for finding and analyzing mask sizes in images
Utility node that automates prompt assembly by replacing placeholder tokens with mapped artist or style names based on ID sequences, supporting multiple prompts and flexible ID handling strategies.
Custom ComfyUI node for generating consistent character images using Ideogram API v3's character reference feature. (Description by CC)
AnimeAPI nodes for ComfyUI: anime image generation at $0.003 per image, no local model or GPU needed.
AntiMatter custom nodes for ComfyUI. Includes aspect ratio tools, image/video folder batch loading and saving, prompt line/JSON scene loaders, LM Studio local API prompt agents, and a text file appender with scene-aware replacement.
Simple UI extension to track Modal GPU credits in real-time.
🚀 ComfyUI Light Pack — turn ComfyUI into a powerful 3D and image-processing toolbox. 40+ practical nodes for mesh repair, UV workflows, PBR previews, texture tools, image management, and workflow control — built to solve real production problems and save hours of manual work.
The nodes Anymatix runs on: URL-addressed model loaders that download what a workflow asks for, plus CLIPSeg, masking, SAM2 coordinates and the save nodes Anymatix reads its results from.
Measure the combined wall-clock time of a section of ComfyUI nodes.
A node pack for cerebras, groq, and SambNova hosted models.
Node pack that loads the official DiffiT checkpoint, creates latent noise, samples with the official pipeline path, and decodes with the selected VAE.
Unofficial EfficientViT_sam_nodes for the https://huggingface.co/mit-han-lab/efficientvit-sam models
This custom node pack provides Marigold IID nodes and PaGeR ERP geometry nodes.
Model Downloader + API-Friendly Wildcard Processor for ComfyUI
A powerful ComfyUI custom node collection for keyframe-based video generation and batch image processing. This extension provides advanced tools for creating smooth video sequences from keyframes with intelligent conditioning and latent space interpolation.
Local ComfyUI nodes for Meta VideoSeal invisible video watermarking.
Four Apex nodes for Krea 2 generation, LoRA control, metadata-aware saves and a resizable image preview.
Apixly AI API for ComfyUI: image generation, reference image edits, video generation, free creative controls and task recovery.
A ComfyUI node to implement Together AI API image generation
Comfyui wrapper of Dreamlight by yongliu20
A ComfyUI custom node for estimating light direction and quality from normal maps using luma masking. The system analyzes surface normals to infer lighting information for downstream tasks like adaptive relighting, directional masking, or stylized effects.
Nano Banana, Seedream, Seedance, Kling, Veo and FLUX in ComfyUI through Liquid Inference, with a USD cap on each run.
ComfyUI plugin that decodes a single image from data and outputs it as IMAGE for use in subsequent nodes. (Description by CC)
Sign-on-arrival save nodes for ArcVelvetOS: ARC Save (image, stable) and ARC Save Audio (WAV, alpha). POSTs encoded media to the ArcVelvet arcIngest API; the C2PA-signed copy comes back and lands in your output directory. Server-side prompt moderation; plaintext prompts never enter the public manifest unless you opt in. Audio node is alpha — audio-content moderation (transcript scanning, voice-clone detection) not yet covered; see README before public use.
NODES: ReadCSVRowNode
Collection personnelle de custom nodes pour ComfyUI (arrssenne).
ComfyUI node for working with complex prompts with support for variables, randomness, and basic logic.
ComfyUI node to make memes and demotivators
Utility ComfyUI custom nodes that add some extra features for your workflows (random int for cli runs, unix timestamp, boolean random, string match switch, etc).
Image prompt engine for ComfyUI: one node turns a brief plus reference images into a finished image prompt via a resident llama.cpp VLM, then encodes it to CONDITIONING. Tabbed panel with a drag-and-drop reference gallery, reference sets, @-insert reference tags, persona identity described once, composite/batch captioning and a prompt-list mode.
Box layouts for MOSS-Image: draw boxes on a canvas to generate a frame where every box holds what you wrote, or to edit only the boxes of an existing picture (outside the boxes stays pixel-identical). Canvas + Sampler, built to sit around an LLM prompter.
Composable workflow units for ComfyUI.
Quickly select corresponding LoRA models, positive and negative prompts based on preset character names. (Description by CC)
ComfyUI Artify custom nodes. Includes a robust image resize node with single-pass scale_by behavior and configurable divisible rounding.
ComfyUI Testing custom nodes for Artify XYZ parameter sweeps and in-node HTML grid viewing.
Browse local pose and outfit references, output images, and compose an editable positive prompt.
Batch-processes every video in a folder through a single-video workflow — no loop nodes or external scripts needed. Advances to the next video each time the prompt is queued, so a whole folder can be processed by queuing N times. Also includes an 'Ensure Even Dimensions' node to prevent odd-resolution encoder crashes with yuv420p/H.264 output.
A power-user toolkit for prompt logic, IO, RNG handling, and LoRA formatting.
Loads universal null lalora plus effi abm models. (Description by CC)
Can load universal LoRA models. (Description by CC)
Lightweight LiteGraph performance enhancement.
Efficient lazy path switch for ComfyUI. Skips execution of the inactive branch entirely using native lazy evaluation. Includes live status display.
ComfyUI custom nodes for NeuTTS on-device TTS with voice cloning
ComfyUI Qwen-Image-Edit-2509 node integration for image editing using OpenAI Image API. (Description by CC)
A tool for removing backgrounds from images using ComfyUI.
ComfyUI's Arcane stylization plugin
All-in-one media preview node for ComfyUI.
A simple custom node for ComfyUI that loads and provides an all-in-one dropdown selector to build a prompt string.
This custom node pack is for deployment validation of ComfyUI + Ascend NPU + INT8 linear weight quantization.
A resolution utility for WanVideo I2V with 20 presets, manual override and automatic rounding.
Label AI-generated images per Art. 50 EU AI Act: visible EU icon/text overlay, XMP metadata (IPTC DigitalSourceType) writer with save, and a metadata inspector incl. C2PA detection.
Nodes:DynamicSampler, MeasuredSampler, ResolveMaskPromise
Pre-flight cost estimator with live balance check for ComfyUI Partner Nodes
These nodes were created to personalize/optimize several ComfyUI nodes for my own use. You can replicate the functionality of most of my nodes by some combination of default ComfyUI nodes and custom nodes from other developers.
A specialized node pack heavily focused on secure image metadata editing and manipulation, supported by a suite of JSON, text, and file system automation tools. Optimized for the V3 API (Nodes 2.0).
ComfyUI custom nodes for CivitAI cloud image generation via API
🎧Ultra High-Quality Voice Cloning, Chinese and English.
NODES: YAML Image Cycler (Full), YAML Image Cycler (Simple), YAML LoRA Extractor, YAML LoRA Loader, YAML LoRA Selector
A custom ComfyUI node that lets you interactively review and remove individual frames from a video batch before it continues through your workflow.
调用kling 中国大陆api,实现kling的使用
ComfyUI nodes for Moebius (hustvl) - 0.22B mask-driven inpainting / object removal, 10B-level quality, no text encoder, pure PyTorch (Blackwell/Ada ready).
crypswolfy69 nodes — Alchemist (any-LLM prompts/captions), Prompt Rotate, and a Hugging Face model browser/downloader for ComfyUI.
use comfyui to image embeddings
Professional COCO-WholeBody 133-keypoint dance animation system for ComfyUI
Save disk space by storing LoRA and LoKr files as INT8: about half the size for bf16/fp16, a quarter for fp32. Quantizer, folder batch and loader nodes. Storage saving only; VRAM use is unchanged.
Small ComfyUI custom nodes for building a Flow-style video continuation workflow with LTX 2.3.
Load 2 to 8 images at once with batch operations. (Description by CC)
A custom node for ComfyUI designed to generate structured prompts for product display images based on analysis results from Vision Language Models.
ComfyUI custom nodes implementing DiffSynth V2 call interface for Qwen-Image-Layered-Control-V2 with layered image decomposition and brush control support. (Description by CC)
A collection of utility nodes for ComfyUI workflows.
NODES: OCR Location Node, Image Erase Node, Chat Overlay Node, Extract JSON Node.
Compare and save unique workflows, count tokens in prompt, and other utility.
7500 prompts at Primeia
Custom nodes for ComfyUI providing advanced model architecture checking and dynamic memory guards, Optimization workflows through smart conditional logic and system resource awareness.
Nodes for images, prompts, video and music. Image: random/sequential loader for any folder, touch gallery picker for the input folder, loader with an on-node crop editor, format detector, booru tag search (5 sites), Ollama anatomy-defect filter. Text: 52prompts.com generators, MiniMax H3 video prompt writer (Ollama or a Qwen/Gemma CLIP, validated), Easy Ollama, text concatenate. Model: LoRA and any-model selectors. Latent: WAN latent start trim. Audio: YuE2 songs, VoxCPM speech and voice cloning.
depricated. new node pack: https://registry.comfy.org/publishers/bartenevve/nodes/comfyui-bart-nodes
A simple math node that performs linear interpolation between a base range and a mapped range.
Simple Linear Scheduling node for ComfyUI. Missing in the custom schedulers.
Some custom nodes for comfyui
Simple nodes to control llamacpp in comfyui.
ComfyUI custom node for sweeping LoRA model strength per latent sample in a single batch.
ComfyUI custom nodes for saving one compact metadata JSON per generated image and viewing embedded image metadata.
This node applies a custom 4x4 color matrix to an image using GPU acceleration via PyTorch.
A high-quality sweep light effect node designed specifically for ComfyUI, supporting multiple parameter adjustments and anti-aliasing.
ComfyUI bridge nodes for TuNan Paint Bridge. Requires the Photoshop plugin (.ccx).
A ComfyUI custom node that renders novel-view previews from image pairs using 3D Gaussian Splatting with sparse scene reconstruction.
A ComfyUI custom node that blends a natural image with a stripe pattern image using pixel-wise multiplication, producing a geometric composite output.
A custom node extension that provides model version tracking, auto-update system, multi-GPU sync, and metadata management for HuggingFace and CivitAI models.
Lightweight custom node that trims a native ComfyUI VIDEO object by timestamps or frame numbers.
An extraction of the one node I actually wanted from ComfyUI_essentials