ComfyUI Custom Extensions

Official implementation in ComfyUI of CVPR 2025 paper 'HyperLoRA: Parameter-Efficient Adaptive Generation for Portrait Synthesis'

A Fast, Accurate, and Detailed Line Detection Preprocessor. Anyline is a ControlNet line preprocessor that accurately extracts object edges, image details, and textual content from most images. Users can input any type of image to quickly obtain line drawings with clear edges, sufficient detail preservation, and high fidelity text, which are then used as input for conditional generation in Stable Diffusion.

ComfyUI node for distributing video latent frames across prompt segments with flexible inline and block syntax for timing control. (Description by CC)

This is a low-dependency node pack primarily dealing with masks. The author recommends using Impact-Pack instead (unless you specifically have trouble installing dependencies).

ComfyUI nodes to use GIMM-VFI frame interpolation

Recursive MiniMax H3 scene loops with motion and audio continuity, visual planning, review gates, atomic checkpoints, resume, and final video assembly.

The ComfyUI version of [a/LivePortrait](https://github.com/KwaiVGI/LivePortrait).

Scheduled prompts, scheduled float/int values and wave function nodes for animations and utility. compatable with [a/framesync](https://www.framesync.xyz/) and [a/keyframe-string-generator](https://www.chigozie.co.uk/keyframe-string-generator/) for audio synced animations in Comfyui.

The implementation for PuLID-Flux, support TeaCache, no model pollution.

A custom ComfyUI node for generating images using the HiDream AI model. Uses quantization for lower memory usage.

nodes for flux ipadapter released by shakker-labs

This extension integrates [a/MV-Adapter](https://github.com/huanngzh/MV-Adapter) into ComfyUI, allowing users to generate multi-view consistent images from text prompts or single images directly within the ComfyUI interface.

Recursive MiniMax H3 scene loops with motion and audio continuity, visual planning, review gates, atomic checkpoints, resume, and final video assembly.

A ComfyUI custom node that simply integrates the [a/OOTDiffusion](https://github.com/levihsu/OOTDiffusion) functionality.

The ultimate solution for managing image metadata and multi-tool compatibility. ComfyUI node version of the SD Prompt Reader.

ComfyUI nodes to use Depth Anything V3 - the latest depth estimation model from ByteDance. For now monocular depth, camera pose estimation and 3D point clouds/gaussians are supported. Models autodownload from HuggingFace (depth-anything org). This is a first draft, let me know if you have any feedback! :)

DeepFuze is a state-of-the-art deep learning tool that seamlessly integrates with ComfyUI to revolutionize facial transformations, lipsyncing, video generation, voice cloning, face swapping, and lipsync translation. Leveraging advanced algorithms, DeepFuze enables users to combine audio and video with unparalleled realism, ensuring perfectly synchronized facial movements. This innovative solution is ideal for content creators, animators, developers, and anyone seeking to elevate their video editing projects with sophisticated AI-driven features.

Enable tensor split multi gpu capability in Comfy UI

Automate calculation depending on image sizes or something you want.

ComfyUI IF Trellis creates a 3d mesh from a single view or multi angle pictures

Refine and improve the quality of small faces in MiniMax H3 video: per-frame face tracking, normalised crops, AV-latent injection and feathered stitch-back.

This node manipulates the pixel art image in ways that it should look pixel perfect (downscales, changes palette, upscales etc.).

A simple node that can dynamically adjust the reserved memory of a workflow in real-time, used to avoid the utilization of shared memory.

Nodes to use Omini Kontext framework for multi-image reference using Flux Kontext model. It has node to for Nunchaku compatibility. The is an editor node to place the character/product on the image for spatial control.

ComfyUI implementation of [a/Omost](https://github.com/lllyasviel/Omost), and everything about regional prompt. NOTE: You need to install ComfyUI_densediffusion to use this node.

Nodes: Random Prompts, Combinatorial Prompts, I'm Feeling Lucky, Magic Prompt, Jinja2 Templates. ComfyUI-DynamicPrompts is a custom nodes library that integrates into your existing ComfyUI Library. It provides nodes that enable the use of Dynamic Prompts in your ComfyUI.

Advanced CLIP Text Encode (if you need A1111 like prompt. you need this. But Cutoff node includes this feature, already.)

ComfyUI Custom Node for HeartMuLa AI Music Generation and Transcript Text

ComfyUI nodes to use [a/DepthAnythingV2](https://depth-anything-v2.github.io/) NOTE:Models autodownload to ComfyUI/models/depthanything from [a/https://huggingface.co/Kijai/DepthAnythingV2-safetensors/tree/main](https://huggingface.co/Kijai/DepthAnythingV2-safetensors/tree/main)

Nodes for convenient prompt editing, making many common operations prompt-controllable

My own version 'from scratch' of a self-rescaling CFG. It isn't much but it's honest work. TLDR: set your CFG at 8 to try it. No burned images and artifacts anymore. CFG is also a bit more sensitive because it's a proportion around 8. Low scale like 4 also gives really nice results since your CFG is not the CFG anymore. Also in general even with relatively low settings it seems to improve the quality.

[a/MusePose](https://github.com/TMElyralab/MusePose) is an image-to-video generation framework for virtual human under control signal such as pose. NOTE: You need to download weigths manually from: [a/https://huggingface.co/TMElyralab/MusePose](https://huggingface.co/TMElyralab/MusePose).[w/The repository name has changed. If you are not receiving updates, please delete the existing node and reinstall it.]

Warning, uses experimental package `comfy-env` to attempt a one click isolated install. Will download and use pixi package manager. Automatic skeleton extraction for ComfyUI using UniRig (SIGGRAPH 2025)

Noise Free images with Euler Discrete Scheduler in ComfyUI with Z-Image or other models

Custom nodes for the paper: Chord: Chain of Rendering Decomposition for PBR Material Estimation from Generated Texture Images

Seamless video connection and multi-reference image support for Wan Animate and Scail-2.

This extension contains a tiled sampler for ComfyUI. It allows for denoising larger images by splitting it up into smaller tiles and denoising these. It tries to minimize any seams for showing up in the end result by gradually denoising all tiles one step at the time and randomizing tile positions for every step.

Creative suite for ComfyUI: fullscreen Image Composer, Paint, 3D Builder, Image Crop and AudioReact editors, plus Load/Save Image, Save Mp4, Load Video, Inpaint Crop and Stitch, Outpaint, XY Plot, Prompt library, LoRA Loader, Control Panel, Notes, Labels, and canvas tools for aligning and grouping nodes.

A ComfyUI custom node wrapping FLUX.2 [klein] workflow into a single UI widget with text-to-image, image-to-image, editing, inpainting, faceswap, pose, and upscaling modes.

Nodes: Download the weights of DragNUWA [a/drag_nuwa_svd.pth](https://drive.google.com/file/d/1Z4JOley0SJCb35kFF4PCc6N6P1ftfX4i/view) and put it to ComfyUI/models/checkpoints/drag_nuwa_svd.pth [w/Due to changes in the torch package and versions of many other packages, it may disrupt your installation environment.]

Improve the interactive experience of using ComfyUI, such as making the loading of ComfyUI models more intuitive and making it easier to create model thumbnails

Extension for Automatic1111 and ComfyUI to automatically create masks for Background/Hair/Body/Face/Clothes in Img2Img

Better version for BiRefNet in ComfyUI | Both img & video

ComfyUI custom node implementation of [a/ProPainter](https://github.com/sczhou/ProPainter) framework for video inpainting.

A ComfyUI extension that provides nodes for memory cleanup, including VRAM and RAM cleanup functions to optimize ComfyUI performance during long running workflows.

Webcam, MIDI, Spout, and GLSL support with animation via tick. Features wave-based parameter modulation, math operations, universal value conversion, shape masking, image channel ops, batch processing, dynamic bus routing, GIPHY and SPOUT integration. Load images/videos from URLs, save output anywhere, and apply transformations like flattening, cropping, and color adjustments. Includes tools for color blindness simulation, stereograms, and stereoscopic imaging—plus much more!

FlashVSR: Towards Real-Time Diffusion-Based Streaming Video Super-Resolution,this node ,you can use it in comfyUI

These custom nodes provides features that allow for better control over the effects of the text prompt.

a comfyui custom node for [a/MimicMotion](https://github.com/Tencent/MimicMotion)

image to prompt by vikhyatk/moondream1

KayTool nodes is designed to enhance the efficiency of building ComfyUI workflows. It includes a variety of practical nodes: support for efficient models like BiRefNet and RemBG for background removal and mask post-processing, wireless data transfer (Set & Get ), AI translation (Tencent and Baidu), dynamic mathematical operations, image processing (size extraction, color adjustment, background removal, mask blurring and expansion), flexible text handling, precision sliders, advanced image saving with metadata support, and more.

This node pack provides nodes that complement the Impact Pack, such as the UltralyticsDetectorProvider.

Unofficial implementation of APISR for ComfyUI, both image and video

ComfyUI implementation for [a/ELLA](https://github.com/TencentQQGYLab/ELLA).

ComfyUI nodes that support video generation by start and end frames

ComfyUI Wrapper for [a/Hunyuan3D v2.1](https://github.com/Tencent-Hunyuan/Hunyuan3D-2.1) - From Images to High-Fidelity 3D Assets with Production-Ready PBR Material

A ComfyUI custom node for creating fixed-frame MP4 reference videos from multiple subject images and a background image, designed for LTX 2.3 MSR (Multiple-Subject-Reference) LoRA workflows.

Nodes for SeC (Segment Concept) - SOTA video object segmentation. Uses SeC-4B model by OpenIXCLab. Also includes coordinate plotter for visualization

This project is used to enable [a/ToonCrafter](https://github.com/ToonCrafter/ToonCrafter) to be used in ComfyUI. You can use it to achieve generative keyframe animation And use it in Blender for animation rendering and prediction

This repository wraps the latest BiRefNet model as ComfyUI nodes. Compared to the previous model, the latest model offers higher and better matting accuracy.

Easily use Stable Video Diffusion inside ComfyUI!

This repository is the official implementation of the [a/HelloMeme](https://arxiv.org/pdf/2410.22901) ComfyUI interface, featuring both image and video generation functionalities. Example workflow files can be found in the ComfyUI_HelloMeme/workflows directory. Test images and videos are saved in the ComfyUI_HelloMeme/examples directory. Below are screenshots of the interfaces for image and video generation. NOTE: 'HelloMeme: Integrating Spatial Knitting Attentions to Embed High-Level and Fidelity-Rich Conditions in Diffusion Models'

[a/CatVTON](https://github.com/Zheng-Chong/CatVTON) warpper for ComfyUI

This is a toolset designed for ComfyUI by LAOGOU-666, providing a series of practical image processing and operation nodes, making our operation more intuitive and convenient

A set of ComfyUI nodes designed and fine-tuned specifically for the Z-Image model. Pushing the best image generation model to its limits!

Implementation of DepthFlow nodes for ComfyUI, adds a 2.5D parallax effect to images and videos. Compatible with Ryan's Flex system.

SHARP integration for ComfyUI - Monocular 3D Gaussian Splatting in under 1 second. Generate 3D Gaussians from a single image using Apple's SHARP model.

NODES: Comfly_Mj, Comfly_mjstyle, Comfly_upload, Comfly_Mju, Comfly_Mjv, Comfly_kling_videoPreview

Wrapper for PyramidFlow -models: [a/https://github.com/jy0205/Pyramid-Flow](https://github.com/jy0205/Pyramid-Flow)

A comprehensive set of custom nodes for ComfyUI, focusing on utilities for image processing, JSON manipulation, model operations and working with object via URLs

The ComfyUI Web Viewer by [a/vrch.ai](https://vrch.ai) is a custom node collection offering a real-time AI-generated interactive art framework. This utility integrates realtime streaming into ComfyUI workflows, supporting keyboard control nodes, OSC control nodes, sound input nodes, and more. Accessible from any device with a web browser, it enables real time interaction with AI-generated content, making it ideal for interactive visual projects and enhancing ComfyUI workflows with efficient content management and display.

Implementation of AutoGen inside ComfyUI. This repository is under development, and not everything is functioning correctly yet.

Custom nodes for using [a/Tripo](https://www.tripo3d.ai/) in ComfyUI to create 3D from text and image prompts.

Photoshop-like layered canvas editor to your ComfyUI workflow. This node is perfect for complex compositing, inpainting, and outpainting, featuring multi-layer support, masking, blend modes, and precise transformations. Includes optional AI-powered background removal for streamlined image editing.

All the tools you need to save images with their generation metadata on ComfyUI. Compatible with Civitai & Prompthero geninfo auto-detection. Works with png, jpeg and webp.

Standalone ComfyUI custom node for Nunchaku QwenImage LoRA loading functionality.

ComfyUI Preset Manager, supporting various preset templates and workflow management

Nodes:MZ_Flux1PartialLoad_Patch. Tool nodes related to flux1

Krea 2 edit (reference image) nodes for LoRAs trained with ai-toolkit. Encodes prompts + reference images through the Krea 2 Qwen3-VL text encoder and patches the model to consume reference latents, with optional KV-cache conditioning.

ComfyUI nodes for anime prompt selection, random prompt generation, background and character tag composition, and Anima LoRA search, download, and loading.

This is a copy of [a/facerestore custom node](https://civitai.com/models/24690/comfyui-facerestore-node) with a bit of a change to support CodeFormer Fidelity parameter. These ComfyUI nodes can be used to restore faces in images similar to the face restore option in AUTOMATIC1111 webui. NOTE: To use this node, you need to download the face restoration model and face detection model from the 'Install models' menu.

SAM 3D Body integration for ComfyUI - Robust full-body human mesh recovery from single images. Reconstruct 3D human meshes with pose, shape, and hand details. First draft, please open a github issue if you have any problems or feature requests :)

A comprehensive ComfyUI extension suite encompassing both front-end interaction and low-level memory scheduling, designed to help beginners master workflows and simplify local deployment.

All-in-one MiniMax H3 video generation for ComfyUI: T2V, I2V, R2V, audio drive, custom keyframes, video extension and Motion Context chains in a single node.

A ComfyUI custom node that integrates Google's Gemini Flash 2.0 Experimental model, enabling multimodal analysis of text, images, video frames, and audio directly within ComfyUI workflows.

Enhance MiniMax H3 and Seedance 2.0 video prompts, Music 3 and YuE2 music text, and Qwen Image 2.1 image prompts with cloud or local GGUF LLM providers.

Nodes such as CLIP Text Encode++ to achieve identical embeddings from stable-diffusion-webui for ComfyUI.

Custom node to convert the lantents between SDXL and SD v1.5 directly without the VAE decoding/encoding step.

An extension for ComfyUI that allows hot reloading. Once installed, you can preview changes in real-time while developing custom nodes or installing plugins without restarting ComfyUI.

This extension contains 6 nodes for ComfyUI that allows for more control and flexibility over the noise.

Joy Caption is a ComfyUI custom node powered by the LLaVA model for efficient, stylized image captioning. Caption Tools nodes handle batch image processing and automatic separation of caption text.

ComfyUI nodes for SCAIL input processing

Nodes: load Image with metadata, get config data, load image from base64 string, Load Loras From Prompt, Generate Latent Noise, Combine Two Latents Into Batch, General Purpose Controlnet Unit, ControlNet Script, Content Mask Latent, Auto-Photoshop-SD Seed, Expand and Blur the Mask

The project integrates the Retrieval Augmented Generation (RAG) tool [a/Llama-Index](https://www.llamaindex.ai/), [a/Microsoft's AutoGen](https://microsoft.github.io/autogen/), and [a/LlaVA-Next](https://github.com/LLaVA-VL/LLaVA-NeXT) with ComfyUI's adaptable node interface, enhancing the functionality and user experience of the platform.

ComfyUI implemtation for NAG

Bilateral Reference Network achieves SOTA result in multi Salient Object Segmentation dataset, this repo pack BiRefNet as ComfyUI nodes, and make this SOTA model easier use for everyone.

A set of useful nodes for convenient use of ComfyUI, including: Seed randomization before the generation process starts, with saving of the last used values and the ability to automatically interrupt the current generation; A function to pause the generation process; Slider nodes for convenient control of input parameters; An alternative version of the standard Reroute node.