ComfyUI Custom Extensions
Browse 9,446 ComfyUI extensions and 73,215 custom nodes.
набор нод: google translate, Concat Images, Flux/SD Latent Image
A ComfyUI custom node for text-to-speech integration with a local TTS server.
A custom node that wraps DeepSeek-OCR as a ComfyUI plugin, providing powerful OCR recognition and document parsing capabilities.
This is a ComfyUI node used to resize input images to a resolution suitable for the qwen-image-edit model.
This extension collapses 'future warning' messages in your Console
A ComfyUI custom node for baking LoRAs into base models and extracting LoRAs from merged/finetuned models. This node provides advanced functionality for managing LoRA weights with support for format conversion between sd-scripts and ComfyUI formats.
A ComfyUI custom node suite for profiling, visualizing and steering attention heads in LTX-Video 2.3 (distill & dev).
This repository contains custom ComfyUI nodes designed to integrate with [a/DreamWaltz-G](https://github.com/Yukun-Huang/DreamWaltz-G), a cutting-edge model for generating expressive 3D Gaussian avatars using skeleton-guided 2D diffusion.
Enhanced LatentSync lip-sync node: no-face-safe, loads models from ComfyUI checkpoints, OOM-guard chunking for 24 GB VRAM.
UX Tools to enhance ComfyUI
A UI suite for ComfyUI including Focus mode and layout enhancements.
Load images with automatic prompt extraction from Civitai URLs, caption files, or EXIF metadata. Features smart dataset detection and dynamic preview updates.
FlashVSR+ video and image upscaling for ComfyUI, with optional spatial DiT tiling for large outputs and limited VRAM. Set tile width, height, overlap, and cosine, linear, or Gaussian blending. Includes streamed VIDEO processing with audio/FPS preservation, optional tiled VAE decoding, and INT8 ConvRot. NVIDIA CUDA required.
A fast drop-in replacement for ComfyUI's Upscale Image (using Model). Batches frames, blends tiles on the GPU, and defaults to no tiling.
A super simple node that outputs common video resolutions as 2 integers for Hunyuan and others!
A custom node for ComfyUI that provides a real-time 7-segment LED display timer to track execution duration and progress during batch image generation.
ComfyUI KSampler with preset save/load/delete
Run AlibabaResearch SparkDiffusion (RoLA sparse attention + CrossDistill few-step) Wan2.1/2.2 video generation from ComfyUI in an isolated runtime, with RTX 50-series / FP8 focused support. Independent community integration, not an official AlibabaResearch project.
HWP custom node pack for ComfyUI — global seed control and workflow utilities
High-performance Core ML diffusion model converter
Various custom nodes will be added.
Japanese speech generation with Irodori TTS, ordered reference audio, and audio export for ComfyUI.
Variables Board is a floating control center for ComfyUI workflows that centralizes parameters into a customizable overlay.
Custom nodes to aid in the exploration of Latent Space
Implements proper multitouch zooming and panning into ComfyUI to make it more usable on mobile devices.
Implements proper touchpad/trackpad zooming and panning into ComfyUI to make it more usable on laptops.
Custom node that allows storing and accessing strings, meant to be parts of a prompt, in a simplified manner. Partially supports dynamic prompt syntax.
Easy deploy API for ComfyUI.
Helper nodes for [a/RunComfy](https://www.runcomfy.com)
Turn your song into a music video, one clip at a time: the IronLeo78 MTV node (MiniMax H3) with its Casting and Direction and Cutting Room modules, Post-Production (refine and upscale). Use with Nodes 2.0.
Upload and compress AI-generated images and enhance your ComfyUI workflows with powerful string utilities and AI integration.
A ComfyUI custom node that directly merges Krea 2 model standards and saves them as needed. Supports deferred lazy-merge patching with optional quantization. (Description by CC)
Dynamic position extrapolation for MOSS-Image: render well past the model's 4 MP training range without retraining.
ComfyUI-translate-offline is a custom node for ComfyUI designed to perform offline text translations using ONNX–based translation models from Hugging Face. The project enables integrating multilingual translations into ComfyUI workflows, focusing on text prompts and conditioning for CLIP models. It supports direct translation between configured language pairs as well as pivot translation via English when no direct model exists. It uses lazy model loading to optimize performance and a SQLite-based caching system to avoid repeated translations.
Advanced context management node for WanVideo using Intent-Aware Mixture of Contexts (MoC) system with hybrid memory, multi-modal retrieval, and temporal diversity. (Description by CC)
A custom node for ComfyUI that allows you to easily browse and select images directly inside the node editor.
[a/FitDiT](https://arxiv.org/abs/2411.10499): Advancing the Authentic Garment Details for High-fidelity Virtual Try-onon
A ComfyUI custom node pack providing local LLM chat, text-to-speech (Kokoro TTS + VibeVoice TTS), and AI voice conversion (RVC).
A/B testing and batch parameter comparison nodes for ComfyUI
Modular prompt assembly nodes for ComfyUI — 8-slot, compact 4-slot, and weighted 6-slot with attention syntax
ComfyUI custom node that shows one toggle per unique frame/group name and can bypass all nodes inside matching frames.
Implementation of Instance nodes, Replicate nodes, and standard Save UI to improve the workflow into ComfyUI.
A ComfyUI node that fetches an image from an HTTP URL and returns it as an image tensor. Useful for API-based workflows.
A lightweight ComfyUI monitoring plugin showing memory and GPU usage in real time.
NODES:ImNewNode, ImAppendNode, MergeNode, SetProperties, SaveToDirectory, batchNodes, redirectToNode, SetEvent, ...
A lightweight ComfyUI chat assistant with model channels, attachments, Markdown rendering, image generation, and selection assistant tools.
A custom node for ComfyUI that expands text prompts using the SuperPrompt-v1 T5 model. This node helps generate more detailed and descriptive prompts from simple input text, which can be particularly useful for image generation workflows.
ComfyUI custom node for Google Vertex AI Gemini text, image, and video workflows.
A YOLOv5 face detection project for ComfyUI.
A grab-bag of ComfyUI custom nodes: workflow-control utilities (lazy image routing, a searchable image loader, model/sampler switches, an LTX-safe latent resizer, a wildcard prompt-file loader) plus 5 best-effort MiniMax H3 companion nodes reconstructed to fill gaps left by node types with no public source.
Generate 3D Gaussian splats from images with InfiniSplat. Gaussian splatting / 3DGS for ComfyUI with native 3D preview, PLY export, batches, and optional camera calibration.
Load & save ComfyUI workflows embedded in image metadata — with optional zlib+base64 compression for JPEG/WEBP — and batch convert folders. PNG drops still work natively; JPEG/WEBP with compressed workflows are handled by this extension.
Interactive rectangle tools for ComfyUI: Select, Crop, Mask, Fill.
ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features for ComfyUI
generate empty latent with aspect ratios
Join Multiline text.
A ComfyUI custom node that adds controlled noise injection during the sampling process for enhanced image generation quality and detail.
A collection of ComfyUI nodes that sculpt diffusion models by applying gradient-based modifications to different layers and blocks.
Watermark node pack for video and image workflows. Corner / center / floating / tiled placement, animated intro and outro copyright cards, and text or PNG-logo overlays with adjustable outline and drop shadow. The watermark text can be typed in the node or driven from an upstream node (text_in), which is handy for burning parameter values into comparison renders. Ships a visual panel for picking styles and uploading a logo, with English and Chinese UI. Works on any IMAGE batch, so it drops into MiniMax H3, Wan, LTX, Hunyuan, VHS or AnimateDiff frame pipelines, and a single photo works too. No extra dependencies beyond ComfyUI's bundled torch / numpy / Pillow.
Face-only photo texture projection for static GLBs in ComfyUI, preserving geometry.
ComfyUI custom nodes for IndexTTS-2 zero-shot text-to-speech synthesis with advanced emotion control
ComfyUI node for MOSS-TTSD (Text to Spoken Dialogue generation)
ComfyUI node for MOSS-TTSD (Text to Spoken Dialogue generation)
A custom node for ComfyUI that integrates Qwen 2.1 prompt rewriting for t2i and image editing
ComfyUI nodes for GLM-TTS, a high-quality text-to-speech system supporting zero-shot voice cloning.
A simple ComfyUI node for generating right eye disparity for VR videos
Custom node for ComfyUI that performs color matching between reference and generated images with support for masking.
ComfyUI nodes for image processing.
This is keit's utility nodes.
Interactive 2D rigging pose editor embedded inside a ComfyUI node. Drag joints to pose a figure, control eye direction, and export as IMAGE.
Load PSD files, rig layers with interactive R / MR / SW control points, and composite as IMAGE + MASK. Supports setup/pose modes, model/pose library, background options, and i18n (ja/en/zh).
Loader node for the pruned FLUX.2-klein-9B text encoder (5.1B). Handles the mixed per-layer FFN/GQA shapes and remapped hidden-state taps that the stock CLIPLoader cannot express.
Spatial and Temporal Attention Masking node for MiniMax-H3 models.
ComfyUI nodes for face based filtering plus image batch merging.
Kinburg's ComfyUI nodes. An end-to-end music-video pipeline: write and sing a song with Siren (AceStep, per-section voices), find where every word was actually sung with Echo (karaoke .ass or burnt-in subtitles that follow the performance rather than the plan), cut the edit to its own music with Orpheus, storyboard and render it with Phantas and Morpheus, then write it out with Save Clip and Save Song. Alongside it a local-LLM toolkit: GGUF nodes with on-demand VRAM unloading and vision, a chat node for talking a task through before you generate, a vision LLM judge, and a managed OpenAI-compatible LLM server that steps aside when ComfyUI needs the card. Plus Ouroboros (self-correcting sampler), Chimera (multi-stage sampler), Audio SR, an interactive HTML image-comparison node, a model and settings preset library, and workflow utilities - loops, accumulators, switches, timers.
Civitai-compatible metadata saving, same-name .txt captions, XY Plot, Tiled Sampling, SAM3, Detailer, PixAI Tagger, wildcards, and Dynamic Prompts to simplify ComfyUI wiring, improve reusability, and carry trial results into the next production pass.
Import a Civitai image's full generation setup into ComfyUI: recipe, model downloads, and a ready-to-queue workflow built from vanilla core nodes.
Nodes:BatchImageResizeProcessor, SingleImagePathLoader, SingleImageUrlLoader
Provides automatic long-form video generation nodes for ComfyUI workflows with intelligent frame blending, RIFE interpolation, and streaming H.264 encoding in a single queue. (Description by CC)
Nodes:XL DreamBooth LoRA, S3 Bucket LoRA
Visual opencv node package based on comfy ui
ComfyUI custom node implementation. (Description by CC)
High-utility nodes for ComfyUI with a focus on Flux 1 Dev workflows and Ultimate SD Upscale enhancement loops.
creating a custom node to test how to use Blocknet controlnets in comfyUI using logic from Diffsynth Studio.
This node loads all images from the selected folder and displays them in a grid. The user can click any image to automatically execute the ComfyUI workflow. Additionally, it supports many RAW image formats as well as HEIC formats.
This tool allows you to download models from CivitAI based on a JSON configuration that defines LORA and checkpoint models. It uses token-based authentication to download files from specified URLs and saves them to specified directories. based on CivitAIDownloader
ComfyUI-Azure-Blob-Storage seamlessly integrates with [a/Azure Blob Storage](https://azure.microsoft.com/en-us/products/storage/blobs/) in ComfyUI. This open-source project provides custom nodes for effortless loading and saving of images, videos, and checkpoint models directly from Azure blob containers within the ComfyUI graph interface.
Use image to search similar images from danbooru using various methods. Notice: the optional API key will be saved to metadata if used
A comprehensive collection of utility and API integration nodes for ComfyUI. Includes image manipulation (4-grid split), string utilities, and multiple API integrations: ImgBB upload, HyprLab upload, Google Image API, Luma AI, FLUX Image API, FLUX Kontext API, and Midjourney proxy integration. Features include image/video hosting, AI image generation, and image description capabilities.
A collection of custom nodes for ComfyUI that enable the creation of various ASCII art effects, from static images to complex, colorized typing animations and video conversions.
ComfyUI custom node for removing watermarks from images using a mask editor and auto-downloading AI model. (Description by CC)
Character separation for anime art in ComfyUI: detection, instance masks, exclusive regions, prompt routing and pose extraction — fixes character blending on multi-character images.
A ComfyUI custom node that integrates Google Photos into your workflows. List albums, load images from specific albums, and search photos directly within ComfyUI. Features customizable image loading options, sorting, and efficient caching for seamless integration of your Google Photos library into AI image processing pipelines.
Two lightweight ComfyUI nodes for Gemini 3 preview models: Gemini 3 Pro (Text) and Gemini 3 Pro Image. (Description by CC)
MiniMax H3 audio-driven video: lock audio into the joint AV latent, render any length in one node, score lip sync offline, and transfer a driven mouth onto a stable face.
ComfyUI nodes to use CompareModelWeights
In Flux, the T5 and CLIP in the text branch are weighted separately to regulate the strength of text-side embeddings.
ComfyUI nodes to use MasaCtrl
Integrated model, LoRA, prompt management, and image saving for ComfyUI.