ComfyUI Custom Extensions
Browse 9,434 ComfyUI extensions and 73,215 custom nodes.
DyPE for FLUX. Nunchaku compatible fork.
This is a node to use OpenRouter API from within ComfyUI. It supports both prompt and image+prompt requests (for multimodal LLMs).
InspireMusic ComfyUI Plugin – ComfyUI Integration Plugin for AI Music Generation A ComfyUI node plugin based on Alibaba’s InspireMusic model, supporting text-to-music generation and music continuation features.
A simple wildcard node for ComfyUI. Can also be used a style prompt node.
Vantage DyPE integrates DyPE (Dynamic Position Extrapolation) into ComfyUI, enabling Flux-based transformer models to produce native ultra-high-resolution images (4K, 8K, and beyond) without retraining or external upscaling.
A node to provide recommended aspect ratios for the Qwen model series.
ComfyUI Wrapper for UniMate
A PCB link render mode for ComfyUI: orthogonal routed traces instead of splines. The routing core of cable-management, packaged alone.
Apply AFFINE noise transforms to latent space to improve image quality, especially with light loras.
Webcam & Screen Share Input Nodes with Cropping and Live Preview of the Cropped Stream
Set to run once, and the SD-PPP Photoshop plugin will be automatically updated when there are updates later, completely eliminating the need for manual updates on the Photoshop side.
A collection of custom nodes and UI extensions for ComfyUI, featuring Multi-Switch, Multi-Pipe, Slider, Slider 2D, SDR to HDR, Save Image, Save Video, Save Audio, Combo Selector, Show Any, Load Image From Folder nodes, and UI extensions (Node Execution Time, Progress Bar).
Qwen Image 2.1, storyboard and digital-human production workbenches for ComfyUI.
Universal OpenAI API-compatible gateway node enabling ComfyUI workflows to call any third-party text2img/img2img image generation service with async task management.
Prompt enhancer nodes for ComfyUI: Qwen Image 2.1 prompt expansion (PE-T2I / PE-I2I), MiniMax H3 video prompts, and general image prompts.
Welcome to ComfyUI-SmolLM3, where we bring the magic of Hugging Face's SmolLM3 language models into your ComfyUI workflow! Whether you're crafting stories, generating ideas, or building AI-powered creativity tools, this node pack makes it delightfully simple.
Gaussian Splat PLY loader + preview nodes (Load/Path/Process/Preview) with bundled gsplat viewer
You can use this node to get information about lora. For example trigger words, description and example images.
Static multi-GPU split for large quantized DiTs - transformer blocks are placed across GPUs once at load time and never move during inference, with hybrid residency for long sequences and activation-space LoRA for packed-int8 models.
Fork of WhatDreamsCost-ComfyUI focused on tested LTX Director extension fixes.
ComfyUI-Easy-DotsOCR is a custom node for ComfyUI that provides text extraction via the DotsOCR engine.
This is a custom node for ComfyUI that provides tools to clean, optimize, and format text prompts. It includes features like converting tags, aligning brackets, and applying weights to prompts.
Make Dia avialbe in ComfyUI.
ComfyUI-Qwen-Image is now available in ComfyUI, Qwen-Image is a powerful image generation foundation model capable of complex text rendering and precise image editing.
ComfyUI nodes for StyleStudio model.
MOSS 布局 JSON 导入与可视化边界框编辑:面板手绘深度图/深度图层、背景图裁剪、DWPose 人体部位识别、尺寸映射与按框色绘制。
Amazon Bedrock nodes for ComfyUI
ComfyUI nodes to use Distill-Any-Depth prediction.
ComfyUI nodes including LLM Text Generation for any LLM server (multiple image/video/audio input, preset management, hide API-key, clean VRAM, auto-unload), Merge LoRA Stacks, Image Pixel Comparison, and other nodes for quality-of-life improvements.
A fast, standalone ComfyUI custom node for Lotus & Lotus 2 (Generative Depth and Surface Normal estimation).
Custom nodes that wrap SteadyDancer I2V generation for ComfyUI.
Nodes:tensor_trans_pil, Make Transparent mask, MergeImages, words_generatee, load_PIL image
This custom node allow upscaling an image by a factor using a model.
Generate high-quality SVG graphics from text descriptions and images using OmniSVG in ComfyUI.
A comfyui custom node package used by myself.
Nodes:KolorsAwesomePrompts
spatial tiling for ComfyUI's native TRELLIS.2 shape decoding and texture decoding
ComfyUI-Seed-Nodes is a custom node library that extends the functionality of ComfyUI, offering advanced image loading and pixelation tools.
A set of ComfyUI nodes for GPC Storage access
Custom node suite for ComfyUI designed for advanced audio processing
Unofficial implementation of DeepSeek for ComfyUI
a comfyui custom node for [a/AniTalker](https://github.com/X-LANCE/AniTalker)
Universal prompt optimizer for any ComfyUI model, with R2V bridges for both MiniMax H3 Director and TimelineDirector.
Dictionary-based prompt composition nodes and dictionary-building tools for ComfyUI.
High-quality Text-to-Speech nodes for ComfyUI using Qwen3-TTS models. Supports voice cloning, voice design, and custom voices with Intel Arc XPU optimization.
Extra keybindings for ComfyUI
A JS editor for five-color masks (RGB + Yellow + Pink) that works with any nodes, with three helper nodes — RGBYPLoadImage, RGBYPMaskBridge, and RGBYPMaskToRegularMasks—for convenient RGBYP mask handling.
ComfyUI custom nodes for Telegram bot integration
Image Detection Bypass Utility for circumvention of AI Detection with comprehensive image processing and post-processing features.
Read and Process data from MIDI devices inside of ComfyUI.
Frontend-only workflow productivity extension providing Alt-right-click smart linking, selection-wide smart connection, input swapping, reroute insertion, data-flow arrangement, and bypass controls.
Scene Framer is a simple utility node for framing multiple shots from a single image.
The unofficial pixel-perfect-depth ComfyUI custom node.
LLM chat, vision, prompt enhancement and image captioning nodes for ComfyUI via any OpenAI-compatible server.
Training-free TRELLIS.2 image-to-3D acceleration: forecast the flow-matching velocity on skipped DiT steps (HiCache Hermite / HiCache++ DMD, via hicache-pp) across the sparse-structure and shape-SLaT stages (texture optional). The README reports historical RTX 5090 results but does not establish current-commit GPU validation. Pairs with visualbruno/ComfyUI-Trellis2.
ARD ComfyUI Ardenius include ARD Control Box, ARD Math nodes and other helper nodes to be added in the future. for more info https://ko-fi.com/ardenius.
Dark minimal theme for ComfyUI inspired by Linear, Vercel, and Raycast. Pure black aesthetic, zinc scale, dot grid background, and ghost/outline buttons.
Entry point for HommageTools node collection for ComfyUI. Handles node registration, imports, and logging configuration.
NODES:Caching CLIP Text Encode for FLUX. Random nodes for ComfyUI for various purposes.
Extra-stealth metadata removal node for ComfyUI
Advanced Gemini 3.x multimodal nodes for ComfyUI.
A real-time system resource monitoring plugin designed for ComfyUI, helping you track computer resource usage during the AI image generation process. It can be freely dragged.
Colorist film tools for ComfyUI: 35mm-faithful film grain (per-channel, band-pass, midtone-weighted, added not overlaid, fresh or coherent) plus a companion warm-highlight halation node.
Apple Silicon fork of Continuity: one node for MiniMax H3, LTX 2.5 and the still families on Metal / MPS
A ComfyUI node that quickly fetches the most commonly used positive and negative prompts for Civitai models (Checkpoint / Lora) from the community, helping you effortlessly enhance your own creations!
Prompt camera assistant for adding camera details to prompts.
load qwen diffusers model
Block-sparse (SLA) attention for MiniMax-H3 -- the inference path the lightx2v SLA turbo LoRA was distilled against.
Advanced LoRA loader with per-block weight control for fine-grained influence over different model layers in ComfyUI
Unofficial support for Hugging Face's hosted inference.
A powerful custom node extension for ComfyUI that enables automatic iteration through multiple prompts with corresponding filename generation. Perfect for batch processing different variations like character face angles, poses, or any sequential prompt workflow.
AMD unified memory lip sync for ComfyUI. Native Wav2Lip inference with no subprocesses - finally works on Strix Halo, ROCm, and all AMD APUs. Features sync tuning, edge blending, and safe tensor casting for unified memory systems.
DEPRECATED - succeeded by Context-Anchored Tile Upscale (comfyui-contextanchoredtileupscale), a native rewrite of this pack's best path. Tiled SD upscaling with custom Guider support: use PerpNegGuider for negative prompts at CFG 1 and higher quality upscales. Context Only Overlap lets each tile use already-generated neighbor pixels as context without re-denoising them, greatly improving consistency and removing seams. Masked Upscaling re-diffuses only a masked region and skips untouched tiles, so parts of an image can be upscaled with different models or settings, or an inpainted patch refined to match. The anchor_context option holds unprocessed areas to the original image during sampling, preventing detail drift between sections and blending seams into the real image. Fork of ComfyUI_UltimateSDUpscale.
Given a set of lists, the node adjusts them so that when used as input to another node all the possible argument permutations are computed.
An 'opinionated' Wave Function Collapse implementation with a set of nodes for comfyui
A non real-time ComfyUI implementation of Kyutai TTS
ComfyUI supports over [a/rhymes-ai/Allegro](https://huggingface.co/rhymes-ai/Allegro), which uses text prompt to generate short video in relatively high quality, especially comparing to other open source solutions available for now.
This is a custom node for ComfyUI that provides the user with a dropdown which contains the recommended resolutions for the Qwen Image Model.
Generate randomised, structured physical person descriptions for image-generation prompts. Sex-aware traits, per-category random/allow_list/fixed/off modes, age ranges, body details, and individual output pins. Full node plus Identity, Face, Hair, Body, and Style segment nodes under the Random Person menu category.
A ComfyUI node pack for Gemini image generation on AI Studio and Vertex AI, interactive crop-and-mask editing with list/grid thumbnail image picking, in-editor upload/refresh, aspect-ratio-safe apply workflows, character planning, prompt tools, and image utilities.
ImageTransceiver is a custom node that enables image generating clients to connect directly to ComfyUI, and send those images in near real-time. For example, an image manipulation program like GIMP can connect an image to a workflow in ComfyUI, and every time the image changes in GIMP, the changes are immediately made in the workflow. Cloning
Allows you to pin images to the JSON of your workflow, migrate with JSON, or embed in image's metadata. supports image compression.
Match image/mask sizes
Download youtube videos/playlists
Premium Quad-Modal suite (Image, Video, Text, Audio) for Pollinations.ai. Includes native Settings menu integration, BYOP support, and an automated real-time model list updater.
Generate images from text prompts using Pollinations' AI models for free.
Testbed for [a/DiT(Scalable Diffusion Models with Transformers)](https://github.com/facebookresearch/DiT). [w/None of this code is stable, expect breaking changes if for some reason you want to use this.]
ASASR 4x diffusion super-resolution (FLUX.1-dev dual-LoRA) for ComfyUI, with native fp8/GGUF support. Non-commercial (CC-BY-NC-4.0 weights).
Comprehensive collection of ComfyUI custom nodes: 🦙 Advanced LLM text generation with Llama-CPP (CPU/GPU acceleration), 🌐 Smart multi-language prompt translation (Google/DeepL/Yandex/Baidu), 🌍 20-language interface toggle, 📸 AI-powered Gemini pose analysis, 🎛️ Smart ControlNet management. Perfect unified package for AI artists and creators. Blog: https://note.com/hirodream44
Nodes:Bing Image Grabber, Zephyr chat, Hermes Chat
CivitAI gallery browser for ComfyUI with protected image preview and prompt editor.
Various cozy nodes, made with 💚 by the CozyMantis squad.
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 nodes that ask a System One decision model (Laya locally, Jev, or Cloudflare Clef) a typed question, so workflows can branch on natural-language judgment and auto-QA generated images.
ComfyUI custom nodes for Kling X-Dub single-person audio-driven lip sync.
A comprehensive image gallery extension for ComfyUI that provides advanced image management, metadata editing, NSFW content filtering, real-time file monitoring, and a full-featured file explorer.