ComfyUI Custom Extensions

Nodes: Load External LoRA Model Only

Lightweight ComfyUI custom nodes for multi-LoRA workflows, image comparison, image saving, and STRING preview.

Complete ComfyUI diffusion pipeline with secure model loading, text conditioning, pipe sampling, and verified downloads.

A ComfyUI implementation of [a/EfficientTAM](https://github.com/yformer/EfficientTAM)

A comprehensive AI API integration and prompt processing toolkit for ComfyUI, providing powerful text and image generation capabilities with advanced prompt manipulation tools.

Misc Nodes: ControlNet Selector Node, Load Optional ControlNet Model, Diffusers Selector, Save Image JPG No Meta, Multi Input Variable Rewrite

ComfyUI nodes for DeepStereo, an auto-autostereogram generator

A collection of utility nodes including image upscale,touchpad support and CSV-driven prompt machine for ComfyUI.

Node: Text Stats (Char & Word Count). A lightweight ComfyUI custom node that takes a text string and outputs character and word counts (INT). Ideal for caption preprocessing, prompt analytics, and dataset workflows.

A Virtual Camera Output For ComfyUI. On Windows, it will use the OBS Virtual Camera driver. So make sure you have OBS installed. Then in your other webcam capable applications, such as Google Meet, Teams, Zoom and even OBS itself, you can connect to the OBS Virtual Camera option and see what you are outputting from ComfyUI.

Application for upscaling videos and images with multi-GPU support, offering a user-friendly interface and high-quality results with minimal effort.

This custom node for ComfyUI enables direct integration with Hugging Face Flux models for image generation via API. Users can leverage the power of Flux models like FLUX.1 [schnell] and FLUX.1 [dev] without leaving the ComfyUI environment

Ultimate all-in-one image generation node with built-in face detailer, hires fix, and color correction

A custom node for ComfyUI that uses various LLM APIs (OpenAI, Anthropic, Gemini, DeepSeek, Qwen/DashScope, xAI, etc.) to optimize and expand your image generation prompts.

Drop-in replacement for ComfyUI's standard KSampler that embeds cryptographic proof of authorship directly into generated images through deterministic noise injection into the latent representation.

Unofficial ComfyUI reference implementation of Hybrid Sensitivity Weighted Quantization (HSWQ) for model optimization.

Unofficial reference implementation of Hybrid Sensitivity Weighted Quantization (HSWQ) for ComfyUI, providing calibration and FP8 conversion nodes for model optimization.

This project contains custom nodes for ComfyUI, developed by ShinChven. The nodes in this package extend the functionality of ComfyUI by providing additional features and utilities.

Store text to Key-Values pair json.

This package has created for generate image from generated image and embedded workflow.

Load model generation data from civitai.

A custom node for ComfyUI that integrates OpenAI last T2S capabilities for free

A ComfyUI extension that integrates PixArt-Alpha models directly into ComfyUI with advanced memory management.

SVDResizer is a helper for resizing the source image, according to the sizes enabled in Stable Video Diffusion. The rationale behind the possibility of changing the size of the image in steps between the ranges of 576 and 1024, is the use of the greatest common denominator of these two numbers which is 64. SVD is lenient with resizing that adheres to this rule, so the chance of coherent video that is not the standard size of 576X1024 is greater. It is advisable to keep the value 1024 constant and play with the second size to maintain the stability of the result.

This adds a node that has both the positive and negative prompts as input in one node. You can just add one node and be done with both Positive and Negative prompts, in place of adding two different nodes for them.

SDR-to-HDR conversion with LumaFlux (arXiv 2604.02787): prompt-free, 8 steps, a frozen FLUX.1-dev plus 71M adapter parameters. Loader, converter with video stabilization, SDR preview, and save nodes for 16-bit PQ PNG and 10-bit HEVC HDR video. The paper repository doubles as the node pack.

An easy custom node that makes the some loaders' input as Text instead of file selector. For example, there are many characters in different loras respectively. If you want to generate different characters' pictures, you have to select corresponding lora, and then edit the prompt. It may cost much time. To solve this problem, You can use it with a chrome extension https://github.com/Sieyalixnet/ComfyUI-Prompt-Formatter-Extension that makes the queue prompt easier when you dealing with massive loras and prompt.

Ecological extension of comfyui using Rust language. only supports Linux.

Utilities for Math and Logic operations.

Tagger used to tag image of but not limited to furry art.

Classic Duck Hunt game integrated as a ComfyUI custom node with interactive gameplay.

Video Combine (Custom Path): combine ComfyUI image frames into MP4/WebM/AVI/GIF and save to any folder on any drive, with optional audio, metadata JSON, preview and an embedded-workflow PNG.

A ComfyUI custom node and reference workflow for sending images from ComfyUI to a local LM Studio vision model, receiving an English FLUX-style prompt or image caption.

Queue progress monitor accessory for ComfyUI that displays rendering progress, processing speed, history, and KSampler preview in a single node. (Description by CC)

SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer

Syntax highlighting and other quality of life improvements for ComfyUI.

sno-test-unclaimed-node

Distilled 1-step ResShift super-resolution (×4 or ×2) as a pixel-space IMAGE to IMAGE node. A ResShift student distilled from a 15-step teacher upsamples any RGB image in one stochastic step (scale dropdown). Self-contained vendored ResShift net (no xformers needed on any GPU); student + vq-f4 VQGAN auto-download on first use.

KSampler-compatible Anima EasyControl image conditioning for ComfyUI.

Custom node for LoRA loader with trigger word display functionality, fully local execution. (Description by CC)

Basic utility for testing diffusion model loss across the timestep schedule. Should work with any native models that use ksampler. This could be used for comparing models, testing captions on an image, etc.

gemini and openai in comfyui

MotionCache LTX2.3 Reproducible Version

Motion-aware denoising cache and batched VAE decoder for MiniMax H3 in ComfyUI

full multi-person pose extraction rather than single-person pose

Sound-related nodes for the Qwen3-TTS project

Draw the motion trajectory for wan move

Advanced Image Saver for ComfyUI: PNG/JPG/WebP/AVIF/TIFF via pyvips, quality controls, metadata toggle, {seed}-style filename patterns, and iterate_mode (overwrite scratch + Save to Assets).

A collection of nodes for using Autogen with ComfyUI [a/AutoGen](https://github.com/microsoft/AutoGen): assistant agents, group chats, code executor, etc.

Bit-depth enhancement and banding removal with classical and deep learning methods

A ComfyUI custom node that calculates width and height while maintaining aspect ratio, making it easier to determine image resolutions with specified aspect ratios and longer side values.

TensorRT Upscaler for ComfyUI

Extract color palettes from reference images and assemble/validate/embed Ideogram 4's structured prompt JSON.

NODES:Image-Harmonizer

A ComfyUI custom node based on GFPGAN for face enhancement, capable of restoring and enhancing faces in images and videos

This is a custom node for ComfyUI that allows you to use the KLingAI API directly within the ComfyUI environment. It is developed based on the KLingAI API documentation. For more details, please refer to the official documentation. [a/KLingAI API Documentation](https://app.klingai.com/global/dev/document-api/quickStart/productIntroduction/overview).

This project integrates the ElevenLabs Text-to-Speech API as a custom node for ComfyUI. It enables seamless text-to-speech conversion directly within ComfyUI, providing the generated audio as a PyTorch tensor for immediate playback or further processing. Ideal for workflows requiring high-quality speech synthesis

Native VIDEO nodes for isolated Echo-WM Flash/Causal generation.

This extension provides the StringToolsConcat node, which concatenates multiple texts, and the StringToolsRandomChoice node, which selects one randomly from multiple texts.

ComfyUI custom node and gallery workspace for prompt libraries, searchable image notes, and reusable extraction workflows.

Run local ComfyUI T2I/I2I/T2V/I2V workflows from Telegram via @FilexaAIBot, using your PC resources for image and video generation.

Stable-Layers / Qwen-Image-Layered RGBA layer decomposition nodes for ComfyUI, plus a PEFT-format LoRA loader that actually applies.

This custom node for ComfyUI allows users to input multiline text and select a specific line by its number. The node processes the input and returns the selected line along with its index.

OllamaKiller is a cross-platform utility node for ComfyUI that automatically manages Ollama processes (`ollama.exe` on Windows, `ollama` on macOS/Linux). It helps free up VRAM by terminating Ollama processes after model execution, improving workflow performance and memory management. Features include graceful process termination, detailed status reporting, and seamless text passthrough.

ComfyUI nodes for LLM Structured Outputs with integration for prompting

ComfyUI custom nodes for line art centerline extraction and SVG manipulation with advanced editing and visualization capabilities.

This repository provides a custom node for ComfyUI that allows managing positive and negative prompts in a structured JSON format. The node supports adding, updating, and logging prompts, ensuring seamless integration into your workflow.

This is a custom node to convert image to ascii art string.

ComfyUI nodes for generating and editing images via Nano Banana 2 through Google Vertex AI. (Description by CC)

Advanced Block LoRA merging node for ComfyUI (allows selective LoRA block merging)

Advanced LoRA merging node for ComfyUI (additive, average, sequential)

Multitrack stem mixer node. Load stems directly into the node, add multi effects per track, mix with faders/mute/solo, preview in real time, and produce a final mix on workflow execution.

A suite of custom nodes for ComfyUI specifically designed to assist Anima models in repairing the loss of details in full-body shots caused by insufficient pixel resolution.

This is a custom node for ComfyUI that allows you to use the UCloud API directly in ComfyUI. UCloud is a high-performance AI computing platform providing powerful image generation models and industry-leading inference speeds. It offers a wide range of AI model options, including popular text-to-image and image editing models, ready to use without local GPU setup.

Japanese-assisted tag autocomplete for ComfyUI text areas. Supports Danbooru tags, bundled natural language tags, local CSV tag files, Japanese aliases, popup settings, and optional Danbooru CSV update from Hugging Face.

FLUXNATION is a fused FP8 CUDA kernel for FLUX.1 that replaces the entire SingleStreamBlock forward pass with a single torch.ops call covering modulation dual GEMMs QKV projection RoPE attention gating and residual all in one shot. On top of that it ships with neuromorphic spike attention which scores every attention block by real dot product similarity and kills the ones that dont matter after image structure has formed keeping only the top 45 percent. Step caching then replays the cached output on alternating spike steps for zero attention compute every other step. The result is 30 percent faster generation than stock ComfyUI on an RTX 4090 with no quality loss. Works with every sampler ComfyUI supports at any step count. FP16 ports for 10 series 20 series and 30 series GPUs are included. A must have for anyone running FLUX.1.

ComfyUI custom node pack for SDNQ quantized models with a modular split-node architecture. Forked from EnragedAntelope/comfyui-sdnq and redesigned into dedicated nodes. Provides 6 dedicated nodes: SDNQ Model Loader, SDNQ LoRA Loader, SDNQ VAE Encode, SDNQ Sampler V2 (general models), Flux2 SDNQ Sampler V2 (Flux2-optimized), and Flux2 SDNQ TorchCompile (performance optimization). Enables 50-75% VRAM savings while maintaining quality. Note: Currently focused on FLUX.2 development; other models not tested.

Connect cloud image and video generation APIs to composable ComfyUI workflows.

Velvet Vice custom nodes shared by the LTX 2.3 and LTX 2.5 ComfyUI workflows.

Velvet Vice MiniMax H3 image-to-video custom nodes with Native/GGUF routing, Quality Refine second sampling, persistent prompting, live preview and finishing tools.

Cinematic AI prompt node for LTX video. Three-pass pipeline: Choreography pre-pass, main generation, lint + auto-repair. Supports Windows, Linux, macOS via llama.cpp (Gemma 4 26B).

Implementation of architectural related graph algorithm in ComfyUI.

The ComfyUI-Visionatrix nodes are designed for convenient ComfyUI to [a/Visionatrix](https://github.com/Visionatrix/Visionatrix) workflow support migration, in particular to extract prompt input params (input, textarea, checkbox, select, range, file) to be used in simplified Visionatrix UI.

Nodes:Inspect, Any String, Model From URL

An experimental STAR Video Upscale Wrapper.

A ComfyUI custom node that converts prompts between different AI image generation models.

TTM (Time-to-Move) node for ComfyUI enabling motion-controlled video generation with Wan2.2 models using dual-clock denoising for independent background and object animation control.

Generate optimized MiniMax H3 video prompts using a local Qwen model (via llama-cpp-vlm). Includes 5 nodes: text-to-video prompts, image/video reference prompts, two-stage custom-rule prompting, multi-turn chat, and a 10-slot rule library.

Use a local LM Studio server as the LLM/VLM backend in ComfyUI.

ComfyUI custom node for loading SVDInt4-quantized Wan/Bernini DiT models.

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).

A ComfyUI custom node for enhancing prompts using OpenAI-compatible APIs (like DeepSeek, OpenAI, etc.).

comfyui_imagetostl is a simple node for ComfyUI that converts grayscale images to STL.

A RefDelta-aware ER-SDE, SEEDS, SA-Solver PEC, and SA-Solver PECE sampler family with spatiotemporal stability control for MiniMax-H3 fused checkpoints.

A custom node that enables MonarchAttention for self-attention in ComfyUI model branches by setting model-level override, affecting only the connected model branch rather than patching globally.

A ComfyUI custom node for safely blending large IMAGE tensors in chunks with explicit resize and compute-device controls.

ComfyUI nodes for AMD AMF VideoSR/HQ video upscaling through FFmpeg sr_amf.

ComfyUI node for background removal, implementing InSPyReNet