ComfyUI Custom Extensions

Advanced camera control prompt generator for ComfyUI that reads camera information from 3D nodes and generates precise, professional camera control prompts.

Frontend extension that adds a sidebar for easy viewing of PNG file metadata.

popup preview for comfyui

A curated set of ComfyUI enhancements built on the V3 API with full subgraph support: real-time resource monitor (CPU/RAM/GPU/VRAM/temp), node execution profiler with per-node timing badges, graph auto-layout (Quick/Dagre/ELK), node navigation with Follow Execution, Play Sound, System Notification, Execution Gate (lazy conditional pass-through), and an image loader with recursive subfolder scanning, custom-folder loading, and metadata extraction. A modern, well-organized alternative to ComfyUI-Custom-Scripts and Crystools.

This is a ComfyUI extension that provides additional API endpoints functionality, primarily designed to support Comfy Portal - a modern iOS client application for ComfyUI.

Video transition effects for ComfyUI using WebGL rendering with 100+ transition variants across categories like dissolve, wipe, zoom, blur, glitch, and more.

[a/AniCrafter](https://github.com/MyNiuuu/AniCrafter): Customizing Realistic Human-Centric Animation via Avatar-Background Conditioning in Video Diffusion Models, you can try this methods when use ComfyUI.

A custom ComfyUI node for managing and downloading models from HuggingFace with preset support

Apply 28-block Anima LoRAs to Anima-2.9B (40 blocks) with the correct block-index mapping, at load time and without converting files.

Adjusts a latent image before the VAE decoding step in order to improve your image quality.

Adjusts a latent image before the VAE decoding step in order to improve your image quality.

ComfyUI custom node for FramePack, supporting 1-frame inferences.

Nodes:IP_LAP Node, Video Loader, PreView Video, Combine Audio Video. the comfyui custom node of [a/IP_LAP](https://github.com/Weizhi-Zhong/IP_LAP) to make audio driven videos!

A collection of custom nodes by arcum42. Currently focused on saving metadata in images so that Civitai picks it up, pulling Civitai information, and misc helper nodes.

ReforgeCFG is a ComfyUI node designed to add details to your image. [a/While it already exists in Comfy Core](https://github.com/comfyanonymous/ComfyUI/blob/80a44b97f5cbcb890896e2b9e65d177f1ac6a588/comfy_extras/nodes_model_advanced.py#L258), it lacks timesteps for adjustment.

带时间戳、标点符号，自动语音识别。给视频自动添加字幕。

Automated Speech Recognition for Chinese.

ComfyUI KSampler Matrix Lab is a custom node for comparing multiple sampler and scheduler combinations in one labeled image grid.

Optimal samplers and schedulers for rectified flow models.

ComfyUI node pack with latent color match and image adjust tools, a universal neural-network latent upscaler (SD1.5/SDXL/Flux/Wan2.2), Flux LayerDiffuse transparent image generation, a multi-image aspect ratio composer, and Flux 3 API nodes. Integrates the standalone Latent_Nodes and WAN_NN_Latent_Upscale packs (DRE-tagged to coexist without conflicts).

ComfyUI plugin for submitting workflows to Thinkbox Deadline for distributed rendering. Enables render farm distribution with configurable job settings, batch processing, and automatic seed variation.

This custom node gives you total control over where your models run and where they live in memory. It solves common Out of Memory (OOM) errors, enables multi-GPU workflows, and fixes persistent crashes when trying to run modern workflows (like Flux/SD3) on CPUs or secondary GPUs.

all top-level standalone functions of OpenCV Python cv2 auto-generated from the type definitions

Local LLM session nodes for ComfyUI using GGUF and llama.cpp, supporting Llama, Mistral, Qwen, DeepSeek, GLM, Gemma, Phi, LLaVA and gpt-oss, enabling both user–model chat and model-to-model dialogue without external runtimes like Ollama.

Custom nodes with a touch of extra UX ✨ Client-side LLM chat, mid-workflow image editing, history for primitives, JSON manipulation, analytics nodes, CivitAI metadata fetching and saving... and more!

Adaptive color fix custom node for ComfyUI with stronger edge protection and luminance-aware chroma fusion.

The implementation of the audio generation model stable-audio-open in ComfyUI enables ComfyUI to achieve text-to-audio functionality.

Translate and extend danbooru tags using Danbot models.

Create Deforum-style zoom animations through iterative feedback loop with pan, rotate, LAB color coherence, anti-blur sharpening, and ControlNet sequence support

Transcribe audio and video files in ComfyUI.

ComfyUI node for NovaSR - Ultra-fast audio super resolution (3600x realtime, 50KB model)

A ComfyUI canvas overlay for notes, labels, highlights, and workflow explanations.

You can use stable-audio-open-1.0 in comfyUI

Nodes: CalculateAestheticScore, LoadAesteticModel, AesthetlcScoreSorter, ScoreToNumber. Aesthetic score for ComfyUI

Custom nodes for ComfyUI implementing the csm model for text-to-speech generation.

Multimodal prompt studio for ComfyUI. Write and refine prompts for image, video and audio models using local or API-based LLMs.

A ComfyUI custom node that provides a basic implementation of Hunyuan Image 3.0 using Transformers.

ComfyUI nodes for integrating Claude Code SDK - enables AI-powered code generation, analysis, and assistance within ComfyUI workflows

This is a ComfyUI node that connects with [a/Roboflow workflows](https://roboflow.com/workflows/build). Roboflow hosts hundreds of thousands of open source and custom object detection models.

Wildcard implementation that can be reproduced with workflows.

Multi-GPU device selection for LTXV2 video generation in ComfyUI

Nodes: NoisyLatentPerlin. This allows to create latent spaces filled with perlin-based noise that can actually be used by the samplers.

VoxCPM TTS for ComfyUI - Text-to-speech with voice cloning, voice design, V2 support, and LoRA training

Wan Video Extender extends a short video or a single image into a longer, temporally consistent clip using a VACE based control latent. It runs in multiple extension loops. Each loop can have its own prompt and optional LoRA, so you can evolve the scene step by step without rebuilding the graph.

A custom ComfyUI node that adds support for the **Qwen 3.5 4B** hybrid (Mamba2 + Attention) text encoder for use with the **Anima 2B** diffusion model.

These are nodes and workflows that can facilitate the creation of animations and video compilations.

This extension provides tools to iterate generation results between runs. In general, it's for cycles.

ComfyUI nodes for KugelAudio - Open-source text-to-speech with voice cloning for 24 European languages

This is a custom node to convert only the Diffusion model part or CLIP model part to fp8 in ComfyUI. VAE fp8 conversion is not supported. The advantage of this node is that you do not need to separate unet/clip/vae in advance when converting to fp8, but can use the safetenros files that ComfyUI provides.

A drop-in replacement for ComfyUI's stock ReferenceLatent node, with per-image strength, per-image timestep gating, per-image MediaPipe auto-masks, megapixel cap on attention cost, and 1–4 image inputs in a single node.

ObjectClear:Complete Object Removal via Object-Effect Attention,you can try it in ComfyUI

Pixydust Quantizer is a custom node extension for ComfyUI that allows for the simplified recreation of tile patterns used in 1990s 16-color PC graphics, offering advanced color quantization and palette optimization features.Additionally, it includes a CRT-like effect node that mimics the phosphor glow and slight color bleeding of old CRT displays, adding a nostalgic aesthetic.With version 2.0.0, it now supports batch image processing, enabling efficient handling of video frames and image sequences.

Nodes: ModelMergeByPreset. Merge checkpoint models by preset

Super fast tensorrt performance with accuate pose estimation of dwpose model, giving the detecting threshold control, plus pose image render and pose json format output. Fine control for pose plotting.

A sequential drag-and-drop image queue node for ComfyUI.

NODES: ImageRewardLoader, ImageRewardScore

Adds a configurable folder watcher that auto-converts Comfy metadata into a Civitai-friendly format for automatic resource tagging when you upload images. Oh, and it makes your UI awesome, too. 💜

Story / sequence image generation for ComfyUI: generate a whole consistent multi-frame story from one JSON prompt in a single node, plus single-image text-to-image / image-to-image, an in-node inpainting studio, a folder-based LoRA dataset + captioning pipeline, and small JSON & type helpers.

This is a custom node collection for ComfyUI that provides some utility nodes.

Collection of image processing utility nodes for ComfyUI including intelligent cropping, aspect ratio scaling, and crop box restoration capabilities. (Description by CC)

Simple custom node pack for nodes I use in my workflows. Includes Dilate Mask Linear for animating masks.

A Text To Speech node using Kokoro TTS in ComfyUI. Supports 8 languages and 150 voices

Lightweight post-processing node for Qwen3-4B conditioning in Z-Image Turbo workflows that refines conditioning with per-token normalization, optional self-attention, and MLP refinement. (Description by CC)

A set of custom nodes for ComfyUI providing advanced image processing, file handling, and utility functions.

The next-generation security, governance, permissions, and multi‑user control system for ComfyUI.

One-node studio UI package for Z-Image Turbo, Flux.2 Klein, Qwen Image Edit 2511, and Krea 2 workflows, including T2I, I2I, inpaint, outpaint, edit, faceswap, angle, and upscale modes.

This repository is a collection of ComfyUI nodes and workflows that can facilitate the creation of animations and video compilations. It will be updated from time to time.

A collection of utility nodes for lora operations in ComfyUI.

ComfyUI Custom Nodes for Inference.Core

Dimensional Latent Perlin is a custom node for ComfyUI that generates Perlin noise in the latent space. This node is designed to work seamlessly with various diffusion models and can be used as an alternative or complement to standard random noise generators in image generation pipelines.

Add nodes that generates danbooru tags by [a/Dart(Danbooru Tags Transformer)](https://huggingface.co/p1atdev/dart-v1-sft).

ComfyUI node for generating animated dotted waveform visualizations with multiple animation styles including teardrop bars

This package contains three nodes to help you compute optical flow between pairs of images, usually adjacent frames in a video, visualize the flow, and apply the flow to another image of the same dimensions. Most of the code is from Deforum, so this is released under the same license (MIT).

FLOAT Talking Portrait (image+audio to talking video) Optimize nodes. FLOAT is the Generative Motion Latent Flow Matching for Audio-driven Talking Portrait. It can animate an image using an audio. Nodes from simple to complex are provided. Optimized for low VRAM and supporting the use of one single file model. You can also experiment with very advanced nodes and dynamic emotions.

A custom node extension for ComfyUI that integrates Google's Veo 2 text-to-video generation capabilities.

One-click tidy-by-role for any ComfyUI workflow — loaders, encoders, samplers, decoders, and outputs each sorted into their own column with coloured group cards and connections preserved. 1200+ node classes pre-classified out of the box.

This is the comfyui implementation of [a/PartPacker](https://github.com/NVlabs/PartPacker): Efficient Part-level 3D Object Generation via Dual Volume Packing.Max varm12G

Multimodal region-specific refinement node for perfect local details, based on RefineAnything with support for color alignment and RGBA image saving. (Description by CC)

ComfyUI-LexTools is a Python-based image processing and analysis toolkit that uses machine learning models for semantic image segmentation, image scoring, and image captioning.

One-click MiniMax H3 long-video generation node: chunked inference, inter-chunk keyframe continuation anchoring, and timeline-based prompt slicing for seamless long videos on limited VRAM.

This extension is a reimagined version based on the [a/ComfyUI-QualityOfLifeSuit_Omar92](https://github.com/omar92/ComfyUI-QualityOfLifeSuit_Omar92) extension, and it supports integration with ChatGPT through the new OpenAI API. NOTE: See detailed installation instructions on the [a/repository](https://github.com/vienteck/ComfyUI-Chat-GPT-Integration).

Custom ComfyUI node that expands structured JSON text into standard prompts for use with models like Gemini, Flux, OpenAI, and Sora. (Description by CC)

This project provides an experimental Tensorrt implementation for ultra fast face restoration inside ComfyUI. Note: This project doesn't do pre/post processing. It only works on cropped faces for now.

BitDance custom nodes for ComfyUI with unified loader, text encode, sampler, and VAE nodes.

Floating chat assistant in the ComfyUI graph editor: reads/edits the active workflow, local docs knowledge base with semantic search, memory, vision, web search, and a workflow-task test suite.

A collection of switch nodes for ComfyUI

Text To Speech (TTS) for ComfyUI

NODES: Memory Cleaner (Force), Memory Status

A custom node extension that hides image previews until you hover over the node.

Advanced nodes optimized for ACE-Step audio generation in ComfyUI.

A tool that organizes workflow nodes into clean, user-friendly layouts

ComfyUI Hunyuan3D-1-wrapper is a custom node that allows you to run [a/Tencent/Hunyuan3D-1](https://github.com/Tencent/Hunyuan3D-1) in ComfyUI as a wrapper.

A custom node extension for ComfyUI that allows mixing multiple models during the sampling process for enhanced image generation.

A ComfyUI custom node for image compression that supports multiple compression formats and parameter adjustments.

ComfyUI-Montagen is an extension designed to enhance the video editing ability of ComfyUI with custom nodes and Montagen. It offers basic video editing features and integrates media clips with the ComfyUI workflow. Furthermore, this extension unlocks the full potential of AI-based video generation and task automation.

Nodes:story_sampler_simple, text2, kosmos2_sampler. I created a dataset for generating short stories [a/Short-Story](https://huggingface.co/datasets/oztrkoguz/Short-Story) and used it to fine-tune my own model using Phi-3.

An advanced image stitching node for ComfyUI.

A ComfyUI custom node that translates text (including prompts) using Google's open-weight TranslateGemma models.

Custom ComfyUI nodes for downloading, converting, and previewing audio/video from YouTube and 1,000+ other platforms

Audio separation (aka demixing) nodes, for Vocals, Instruments, Bass, Drums and Others (experimental Piano and Guitar). Using MDX-Net and Demucs, no extra dependencies, support for batch and resample. Choose between High Quality and Speed. All safetensor models (No ONNX, No PyTorch)