ComfyUI Custom Extensions

Two lightweight ComfyUI nodes for Gemini 3 preview models: Gemini 3 Pro (Text) and Gemini 3 Pro Image. (Description by CC)

将小说章节转换为影视分镜脚本的ComfyUI插件

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

Dynamically balance weights of multiple LoRAs over generation steps for ComfyUI.

Feature extension package for ComfyUI-Crystools.

Upload media files into ComfyUI in a simpler way.

Custom node to use Gemini 1.5 and above for Comfyui to generates theme related prompts for image generators

Adds a smart I/O status dock to stacked nodes for cleaner layouts.

ComfyUI LC Vision Nodes by lonecatone23 — a Qwen-VL GGUF vision/captioning pack built to actually survive shared-VRAM ComfyUI workflows: a Loader + Run split (so the model stays resident instead of being torn down every call), multi-reference input (several independent images plus a video in one call, not just a single batched slot), and an install script that resolves a matching llama-cpp-python vision wheel for the user's own machine instead of pointing at a manual wheel-matching guide. Companion to ComfyUI_LC123_nodes; separate Registry name.

Egregora: Divide & Enhance is a small suite of custom nodes that help you split, enhance, and recombine images, plus a clean SDXL prompt mixer that keeps things simple while staying robust with lot´s of customization.

Creates spatial echo and ambient effects by applying reverb-like processing directly in latent space

Photoshop-style image layer compositor for ComfyUI. Fixed canvas, dynamic layers, move / rotate / scale.

Prompt808 allows you to analyze images, build style libraries, and generate prompts that match any learned aesthetic.

Multiple utilities for ComfyUI

Node to set v-prediction sampling when using SDXL and other models that may not have the necessary metadata to identify it as a v-prediction model. This node is useful for quantized models since they lack the necessary metadata.

Gemma 4 reference-image prompt nodes for ComfyUI

Node pack designed to save images with metadata supported by Civitai.

Dynamic ComfyUI node that generates AI-friendly outfit descriptions from configurable JSON data. Features gender-specific interfaces, 20+ body parts, dynamic makeup system, bias-free output, and extensible design. Includes support for Ollama vision/LLM integration. Perfect for AI image generation workflows.

Store a CSV of prompts where the style can change for each subject. The CSV node initialises with the column (style) and row (subject) names for easy interpretability.

A node to set workflow settings.

Self-contained ComfyUI nodes for Tongyi Z-Image-Turbo high-quality workflows

Third-party API image generation node collection for ComfyUI supporting OpenAI-compatible chat interfaces and GPT Image with custom API endpoints and model configuration. (Description by CC)

A collection of custom nodes for ComfyUI

Text processing utilities for ComfyUI by masslevel.

MRLN Nodes - a multi-domain collection of custom nodes for ComfyUI.

タグ選択、翻訳、編集、整理をまとめて扱うComfyUI向けプロンプト作業台 / A structured prompt workbench for ComfyUI with tag browsing, translation, editing, and organization.

A collection of custom audio nodes for ComfyUI including MP3 tagging and Lyrics Video generation.

Smart, unified model loaders for ComfyUI that support both standard .safetensors and quantized .gguf formats — no switching nodes required. Includes flexible UNET and CLIP loaders that work across models like SDXL, SD3, Flux, and more.

Pause image generation when GPU temperature exceeds threshold.

This node was designed to help with checkpoint configuration. Fee free to add new checkpoint configurations!

Additional ComfyUI nodes to utilise the Transformers pipeline in a simple and modular way.

An integration between comfy ui and petty paint

BoxBox is a lightweight ComfyUI extension for drawing, cropping, resizing, and reinserting specific regions of an image.

Three simple ComfyUI nodes that show the precision ComfyUI is actually using at runtime.

A simple actionbar for ComfyUI.

A simple node demonstrating how to build custom nodes in Rust!

Militant Hitchhiker's Switchblade Pack is a collection of custom nodes for ComfyUI that provide various multi-function capabilities.

A ComfyUI custom node that emits a held, range-bound integer index with multiple advancement modes (loop, clamp, pingpong, random, shuffle) — useful as a Load Image batch index that advances every N runs.

This node can easily switch between vertical and horizontal values with a single button.

A ComfyUI custom node suite for Audio-to-Audio Schrödinger Bridges (A2SB) with state-of-the-art audio restoration, bandwidth extension, and inpainting optimized for modern NVIDIA GPUs.

An all-in-one ComfyUI face detailing node that combines face detection, inpainting, and advanced sampling in a single node — no complex node chains required. (Description by CC)

MultiDiffusion and SpotDiffusion tiled diffusion for ComfyUI, with Anima (COSMOS) 5D latent support.

Cycle through the number input fields on a node using Tab / Shift+Tab. Classic UI QoL extension.

Zero files. Zero storage. Infinite prompts

Advanced multi-dimensional prompt variation system for ComfyUI with 240 pre-defined modifiers and 19 custom nodes. (Description by CC)

This project provides a collection of custom nodes designed for enhanced audio effects in ComfyUI. With an intuitive pedalboard interface, users can easily integrate and manipulate various audio effects within their workflows.

A ComfyUI custom node package for the OmniGen2 multimodal generation model.

A powerful and versatile custom node for ComfyUI that provides multiple ways to load images into your workflows.

A ComfyUI custom node package for ModelScope API integration, providing seamless access to ModelScope's powerful AI models including Large Language Models (LLM), Vision-Language Models (VLM), and Image Generation capabilities.

A ComfyUI custom node module for managing model presets and configurations. This module automatically detects the model, loads/saves presets (sampler, scheduler, steps, CFG, clip-skip, seed, resolution), and generates/caches preview images per model.

Encyclopedic Studio Prompt & Storyboard Director Suite for ComfyUI — FLUX.2 Klein 8-Layer BFL Formula, LTX-Video 2.5 Multi-Shot Cuts & Dub-It, Voice Design & Prosody, and Multi-Reference Storyboards.

This is a simple Hidream block swap node for ComfyUI native nodes, works by swapping upto 18 blocks to the CPU to reduce VRAM.

MIRROR of 'mmaker/sd-webui-color-enhance' for comfyui | mmaker is currently hosting himself | Backup Repo | ORIGINAL https://git.mmaker.moe/mmaker/sd-webui-color-enhance

A simple text randomizer for ComfyUI that can generate random and surprising results

Nodes: QRNG Node CSV. A node that takes in an array of random numbers from the ANU QRNG API and stores them locally for generating quantum random number noise_seeds in ComfyUI

ComfyUI custom node: LoRA stack with auto trigger-word extraction (civitai + safetensors metadata)

This repository contains a collection of custom nodes for ComfyUI, designed for generating various patterns, optical illusions, and performing related image manipulations. All nodes are categorized under 'illusion' in the ComfyUI menu.

A custom layer editor for ComfyUI with three utilities (drop shadow editor, crop editor, and gradient editor).

A lightweight ComfyUI extension implementing S²-Guidance and Perpo-Guidance, two advanced sampling techniques designed to improve prompt adherence and generation robustness compared to standard CFG.

VFX-oriented nodes for ComfyUI: HDR LogC3 / LogC4 decode, EXR I/O, and an interactive ACEScct color grading panel powered by exr-viewer.

Implementation of GoWithTheFlow, original code at [a/https://github.com/Eyeline-Research/Go-with-the-Flow/](https://github.com/Eyeline-Research/Go-with-the-Flow/) and [a/https://github.com/RyannDaGreat/CommonSource/blob/master/noise_warp.py](https://github.com/RyannDaGreat/CommonSource/blob/master/noise_warp.py)

Implementation of GoWithTheFlow, original code at [a/https://github.com/Eyeline-Research/Go-with-the-Flow/](https://github.com/Eyeline-Research/Go-with-the-Flow/) and [a/https://github.com/RyannDaGreat/CommonSource/blob/master/noise_warp.py](https://github.com/RyannDaGreat/CommonSource/blob/master/noise_warp.py)

Upsamples danbooru tags using Dart models.

Generative View Synthesis with Diffusion Models

Nodes which make it much easier to manage prompts

Automatically organize ComfyUI image and video outputs by subject using local Ollama classification, with metadata preservation and full-library Auto-Sort recovery.

Dub and translate audio/video using Pinch AI

Pirog's Nodes - ComfyUI custom node pack providing enhanced sampling functionality

This collection provides four additional nodes for loading and managing Hunyuan Video LoRAs in ComfyUI

Advanced latent space transformations during diffusion sampling with precise control over timing and distribution.

A custom node on ComfyUI that saves images in AVIF format. Workflow can be loaded from images saved at this node.

Custom ComfyUI nodes for iterative video storytelling workflows.

The Polyhedron Suite for ComfyUI. Group-aware LoRA management: stack 10-25 LoRAs with per-group merge modes (SEQ/CONCAT/DARE), per-row decoupled CLIP strength, Trim & Resolve cleanup switches against multi-LoRA interference, thumbnails, trigger words, Civitai fetch, a dedicated Accelerator engine and a passive Merge Analyzer. Plus media I/O: a pinned-folder Media Loader (images, video, audio, batch modes) and a unified Save node for stills, video and audio. Plus sampling: a classic-wiring Sampler with a live preview decoder and a CLIP Text Encode that shares the Token Counter's count.

This repository provides a custom node for ComfyUI to load diffusion models with musubi block-wise scaling.

FRED's enhanced custom nodes for ComfyUI

THIS NODE IS WIP. ComfyUI wrapper for MIDI-3D: Multi-Instance Diffusion for single image to compositional 3D scene generation.

A tool for calculating audio duration that takes audio input and outputs the audio length as a floating-point number. (Description by CC)

Create A1111-style XYZ plot grids in ComfyUI to test multiple parameters and compare results in organized, labeled comparison grids.

ComfyUI-FooterGallery is a ComfyUI extension that adds an image gallery to the bottom of the user interface.

AI Assistant for ComfyUI with agentic capabilities to interact with workflows.

MCP server + multi-CLI embedded terminals that let AI coding agents see and edit your ComfyUI workflows

Mistral AI API's chat completion endpoint in ComfyUI

Complete aesthetic overhaul for ComfyUI workspace with animated wrappers for core nodes, custom text inputs, and a real-time Matrix Execution Monitor.

**MiniMax H3 SeqAttn brings long-sequence MiniMax-H3 generation to low-VRAM ComfyUI setups.** By streaming long-sequence activations through CPU memory and processing attention and MLPs in fixed-size GPU tiles, **denoising VRAM stays bounded instead of growing with sequence length**. The checked-in 8 GiB validation belongs to version 0.4.0's fused DiT path and retired Qwen offload implementation; it is not a Qwen SeqAttn performance claim.

This custom node allow you to upload result images to imgbb.

A standalone custom node for ComfyUI that provides an interactive tag-based interface to enable/disable trigger words from any LoRA loader or gallery node.

VFX and video utility nodes. Avoiding opening other apps/clearing other caches.

A custom node for ComfyUI that integrates the powerful stable-diffusion.cpp inference engine.

Nodes to use [a/latte](https://github.com/Vchitect/Latte) for text to video generation

Seamlessly join multiple video clips using WAN VACE with one click.

A ComfyUI node that generates tiled image compositions with overlapping regions. This approach creates coherent compositions by using the edges of each tile as seeds for neighboring tiles, resulting in seamless transitions.

ComfyUI node that immediately downloads and launches Genshin Impact.

NODES: Extract Displacement Map Node, Displace Logo

A custom node which helps find which custom nodes are importing a particular library

ComfyUI-SceneGenerator is a ComfyUI plugin used to generate scene preview photos from JSON files. This plugin creates scenes based on the provided JSON configuration and produces two different image outputs: one containing only the products and the other containing both the products and the props.

Review-first checkpoint cycling and handpicking suite for ComfyUI. Auto-cycle checkpoints, review generated images, tag checkpoints, and export cleanup scripts without deleting files immediately.

GraphicsMagick-based direct JPEG saver node for ComfyUI. Saves IMAGE tensors directly as JPEG without intermediate PNG files. JPEG-only, previewless, label-aware, and useful for organized batch/inventory workflows. NOTE: GraphicsMagick must be installed separately and available via GM_PATH or PATH.