ComfyUI Custom Extensions

comfyui custom node for depthflow original depthflow website: [a/https://github.com/BrokenSource/DepthFlow](https://github.com/BrokenSource/DepthFlow) check this for installation: [a/https://brokensrc.dev/get/](https://brokensrc.dev/get/)

Permutes a mask batch to present possible additive combinations. Passing a mask batch (e.g. out of [a/SEGS to Mask Batch](https://github.com/ltdrdata/ComfyUI-Impact-Pack)) will return a new mask batch representing all the possible combinations of the included masks. So, a mask batch with two mask sections, 'A' and 'B', will return a batch containing an empty mask, an empty mask & A, an empty mask & B, and an empty mask & A & B.

A prompt manager on the sidebar for ComfyUI, allowing easy saving and copying of prompts.

Custom node for using Prompt S/R in XY Plot Also includes nodes for listing generic parameters like seed and cfg Easy to manipulate as elements are separated by line breaks Designed for use with the XY Plot custom node qq-nodes-comfyui, but may work with other custom nodes as well

ComfyUI custom nodes for Kling X-Dub single-person audio-driven lip sync.

A curated collection of image/latent creation and transformation custom nodes for ComfyUI. Each node is independently developed and maintained, but packaged together for convenient installation.

An advanced aspect ratio calculator and image scaler with flexible scaling modes and intelligent image handling.

This exporter is a plugin for ComfyUI, which can export tasks for [a/LAizypainter](https://github.com/DimaChaichan/LAizypainter). LAizypainter is a Photoshop plugin with which you can send tasks directly to a Stable Diffusion server. More information about a [a/Task](https://github.com/DimaChaichan/LAizypainter?tab=readme-ov-file#task)

Managed native llama-server nodes for ComfyUI

48-Layers Latent Previewer for Lightrick's LTX-2 model.

A small ComfyUI node pack for using NAFNet models (specifically this one to try to get rid of Qwen/Wan VAE grid artifacts: https://huggingface.co/DraconicDragon/NAFNet-VAE-DeGrid) that output a residual.

Resolution maths for latent-space models, done exactly. Real megapixels (10^6, not 1024^2), patch-grid alignment for Krea 2 / MiniMax H3 / Flux / SDXL, and aspect ratios held exact instead of drifting on rounding.

ComfyUI 肖像大师中文版，这是一个用于生成 prompt 文本的 node, 无其他依赖。在原版基础上完成汉化并添加了一个 ImageSaver 用于保存 prompt 等元信息到 exif 的 node

Nodes:Shadertoy, Shader, ColorChannelOffset.

ComfyUI acceleration extension designed specifically for the 170HX GPU that enhances inference performance by optimizing specific operators based on academic research.

SongGeneration inference plugin for ComfyUI that wraps SongGeneration model inference as custom nodes for audio generation workflows. (Description by CC)

Enhanced Text Encoder modules, acd Custom nodes for ComfyUI, device-select CLIP loaders, providing HDR effects, image saving with prompt metadata.

Custom model APIs for text, vision, image generation and editing. Provider management, model discovery, request templates, authenticated management, public-network access with explicit local-service grants or a strict origin allow-list, and English UI with native ComfyUI localization.

Color match nodes for ComfyUI that color-match target images against reference images with manual or auto-ramped blending, ideal for smooth color transitions between video clips.

A single ComfyUI loader node for MiniMax H3 FL2VA and REF2VA diffusion models, for combined i2i+ref2i inference using ComfyUi MiniMax H3 Image And Reference To Video node.

ComfyUI integration for BAAI's Emu3.5 multimodal models for text-to-image generation and multimodal understanding. (Description by CC)

Unofficial diffusers integration of the official SDNQ pipeline to run in ComfyUI. (Description by CC)

ER VFX Nodepack: film-grade VFX tools for ComfyUI — generative editing (FLUX.2 klein/Kontext/Qwen-Edit), lens flares, defocus/bokeh, glow, grain, camera dirt, color correction with vectorscope, relighting, layer compositing, wedge sequencing and more. GPU live previews on every node.

Track usage statistics for all ComfyUI nodes and packages

A modular character prompt generator for ComfyUI

ComfyUI custom nodes for calling text and vision LLM APIs through multiple provider-compatible request protocols supporting OpenAI, Anthropic, and Gemini formats.

Node:GCP Storage Node. Support google-cloud-storage.

experimental wanvideo comfyui node with a singular goal - visually seamless transitions between context windows

This extension provides various custom nodes. literals, loaders, logic, output, switches

This fork of the official StabilityAI repository contains a number of enhancements and implementations.

MiniMax H3 semantic conditioning, audio frame-rate sync, latent upscaling, tiled sampling, batched VAE decoding, motion continuation and SelfLift sampling.

Convenient, high-speed model downloads directly to your Comfy install with blazing-fast, multi-connection support. No more downloading models to your downloads folder and having to manually move them to the proper folder!

Custom node for ComfyUI that generates JSON prompts from natural language descriptions to improve text-to-image generation quality. Supports local LLM integration via Ollama or Google Gemini API.

A ComfyUI LoRA loader and explorer for Anima models with local LoRA browsing, Civitai Anima search, trigger tools, download progress, reference examples, and version update checks.

A custom node for ComfyUI to have many sliders with interactive dynamic behavior

A custom node for ComfyUI to have many sliders with interactive dynamic behavior. This repository includes all necessary code and assets for seamless integration.

Nodes to interact with the mrv2 player

ComfyUI wrapper of IP-Composer

ComfyUI_imgcanvas At present, I have not used the useful comfyui custom node about layer mixing, and I have written a comfyui runtime automatic pop-up window for layer editing node

Allows making a mirror of options that are on a node, for use creating your own UI

Comfyg Switch is a custom node that dynamically selects model configuration parameters based on the chosen checkpoint. It reads model-specific settings from a JSON file (model_configs.json).

A collection of utility nodes for ComfyUI, including resolution presets for film and photography aspect ratios.

Select one line from the text list with a single click. It dramatically improves the process of switching LoRA or prompts.

Jev text interpretation and judgments for ComfyUI: candidate selection, condition evaluation, scoring, number extraction, and local Skill selection. Includes OpenRouter text generation.

HQH-XPU: Intel Arc A770 optimization bundle. Installing this node auto-deploys TINT4 (onednn hybrid), OmniXPU, XPU-CacheClean, ClipProj (XPU), UniversalIO, H3-Latent-Upscaler and HQH-H3-Pipeline nodes.

Automatically calculates and applies the optimal resolution for Wan 2.2, MiniMax H3, and LTX video generation models based on aspect ratio and model requirements.

ComfyUI nodes primarily for seed and filename generation

Standalone all-in-one ComfyUI image editor node with canvas workflow, featuring text-to-image, image-to-image, inpaint, outpaint, upscale, and drawing tools. (Description by CC)

Custom nodes that organize ComfyUI workflow connections by combining multiple links into single pipeline links. (Description by CC)

A collection of nodes to enhance your experience with ComfyUI.

This custom node provides a memory management utility for ComfyUI. It allows you to delete a specific model (checkpoint, etc.) completely from VRAM and system RAM after use, while passing through any other input type unchanged (IMAGE, LATENT, CLIP, STRING, INT, CONDITIONING, VAE, etc.). This is especially useful for low VRAM & low RAM environments, helping to reduce out-of-memory (OOM) errors in long workflows.

Full-featured Kling AI integration for ComfyUI - Video, Image, Audio generation

Image cropping and Image resizing

A comprehensive ComfyUI plugin that enables seamless integration with Substance 3D Designer workflows through command line automation. This plugin provides custom nodes for cooking .sbs files, rendering .sbsar archives, controlling material parameters, and batch processing Substance materials within ComfyUI workflows.

A custom node for ComfyUI that allows you to render GLSL shaders (Shadertoy-style) directly inside your workflows. (Description by CC)

GPU-powered pro-grade image suite cho retouch chân dung. v0.1.0 alpha: Frequency Separation (N-01) + Sub-Pixel Mask Refiner (N-02). Kornia core, tensor thuần không rời VRAM.

This is a set of custom nodes for calling an image translation API within ComfyUI.

ComfyUI extension with patches and nodes. Patches:Preview device, Extension device, Temperature, Memory estimation, Optimizations, Easy generic inputs, Easy multiple inputs. NODES: Image nodes, Image/Area nodes, Image/Blip nodes, Image/Face nodes, Sampling nodes, Patch nodes, Primitive nodes, Primitive/Conversion nodes, Primitive/Process nodes, Workflow nodes, etc...

TouchDesigner-style blob tracking and computer vision effects for ComfyUI. Simple nodes for bright spot detection, plexus connections, and technical aesthetics in video workflows.

Download models from HuggingFace (no API key) and manage your local ComfyUI models (list, move, delete) from a built-in panel.

Advanced Eye Tracking ControlNet System for ComfyUI - Professional eye-tracking with MediaPipe, 6-DOF Kalman filtering, and WAN 2.1/2.2 compatibility

A high-performance ComfyUI custom node for intelligent image resizing with industry-standard presets and smart orientation detection. Optimized for video workflows and batch processing.

RV-style interactive A/B video review player node for ComfyUI - wipe, side-by-side, difference, frame-accurate transport

Nodes: NudenetDetector

Video frame interpolation (VFI). Retime any clip to any frame rate, turn 16 or 24 fps into smooth 60 fps, or stretch it into slow motion. A RIFE alternative with arbitrary-time conditioning and six inference controls.

ComfyUI custom node that automatically generates image generation prompts using AITuber persona dataset (195 characters). Uses LLM to generate prompts based on character index and atmosphere keywords. (Description by CC)

A prompt generation system that manages relationships between prompt components to maintain logical consistency. Integrates with ComfyUI as a custom node.

A ComfyUI node that utilizes Moviepy to convert and send your images or videos to a webhook endpoint directly from ComfyUI.

Comfyui custom node that modifies image channels with various per-channel operations

Touch-friendly gallery picker for ComfyUI image and path widgets. Adds an inline card-grid LoadImage node, a modal picker over stock LoadImage with Input/Output/Temp tabs, and a filesystem browser for VHS path-loader nodes (single image, image directory, video).

A dedicated ComfyUI node that replaces naive interpolation with a trained neural network to resize Wan video latents.

ComfyUI for [a/LLaSM](https://huggingface.co/spaces/LinkSoul/LLaSM)

ComfyUI for [a/style-aligned](https://github.com/google/style-aligned)

Nodes:Load Video to Images, Image to Canny, ControlNet Model Loader, VEDit Model Loader, VEdit Sampler. [a/https://github.com/SingleZombie/DiffusersExample/tree/main/ReplaceAttn](https://github.com/SingleZombie/DiffusersExample/tree/main/ReplaceAttn)

Multimodal Context IR prompt builder for MiniMax H3 video generation in ComfyUI

Composing prompt variants from the same text pieces with ease: • Build your "library" (dictionary) of named text chunks (sub-strings) to use it across the entire workflow. • Compile these snippets into different prompts in-place - with just one string formatting node. • Freely update the dictionary down the line - get different prompts. • Reference text chunks within each other to build dependent hierarchies of less/more detailed descriptions. • A real life-saver for regional prompting (aka area composition).

One click replacement of smart objects or layers in PSD

ComfyUI多功能工具包 - 图像/视频处理、翻译、提示词生成

Easiest way to use SAM3 in ComfyUI (with image/video support)

Ignore the output cache of specific nodes to save RAM.

Anima character styling console and cross-plugin tag arbiter for ComfyUI: slot contract for clothing / pose / expression / scene, 0-4 NSFW tiering, weighted random outfit rolling, and random casting from a 5981-character pool

Professional AI Image Description Generator Based on Zhipu AI GLM-4V multimodal model, batch generate accurate and detailed descriptions for images in Chinese and English

A custom node collection developed for ComfyUI, offering preset dimensions for Latent, loading LoRA from folders, and integrating multiple commonly used custom nodes.

Nodes:Apply Ref UNet, Ref Sampler, Ref Sampler Custom

A furry-friendly ComfyUI extension allowing the interrogation of e621 tags from images.

ComfyUI custom node: prompt authoring for LTX video - Studio panel, brief builder, wardrobe and reference rails

Remove content inside 'think' tag from reasoning llm

ComfyUI node for GFPGAN. GFPGAN aims at developing a Practical Algorithm for Real-world Face Restoration. It leverages rich and diverse priors encapsulated in a pretrained face GAN (e.g., StyleGAN2) for blind face restoration.

Ambrosinus ToolKit - Streamlined workflow export with transparent backgrounds, professional themes, and smart scaling. Perfect for creating clean, high-resolution workflow documentation and sharing.

A custom ComfyUI node that allows you to combine two LoRAs with adjustable strengths and multiple merging methods.

IndexTTS2 精确停顿控制（[pause:N]）ComfyUI 节点：停顿延长/缩短/插入，平均偏差 13ms，支持批量生成与试听验收。Precise pause control for IndexTTS2 — TTS ComfyUI nodes with waveform-domain pause editing, ±13ms accuracy.

Resolve the conflict issue between the original plugin and the new comfy environment, adapt Nunchaku and kontext, support Chinese and custom fonts, and quickly measure the best parameters after model fine-tuning without the need for complicated node connections.

IndexTTS-2.5 的 ComfyUI 自定义节点

Qwen-TTS nodes for ComfyUI

Inteliweb utilities for ComfyUI: notifications, seed, Set/Get, monitoring, memory, routing, prompts, and text tools.

Multi-GPU acceleration and caching for MiniMax H3 video generation in ComfyUI: Ulysses sequence and head parallelism across 1-8 GPUs, sparse attention inside the split, a pipelined exchange, and text-encoder/VAE caches.

Adds math nodes for FLOAT, CONDITIONING, LATENT, IMAGE, NOISE and AUDIO. Allows usage of math expressions with variety of functions and variables, not limited to inputs. List of those is on github of this extension.

A node layout helper extension for ComfyUI. Assists in arranging and aligning nodes on the canvas without modifying workflow files — all operations are frontend-only.