ComfyUI Custom Extensions
Browse 9,446 ComfyUI extensions and 73,215 custom nodes.
Comfpod's fork of ComfyUI nodes for language and vision models using OpenAI-compatible chat completion APIs.
Batch-capable GPU layerstyle nodes for ComfyUI.
Four-view Mixture of Diffusers sampling for SDXL in ComfyUI.
A set of custom nodes for ComfyUI that allows you to manage, save, and load system prompts (roles) easily.
ComfyUI custom node package. This custom node features multiple practical functions, including global variables, flow control, obtaining image or mask dimensions, and Dominant Axis Scale.
An extension (custom node) for Comfy Desktop that adds a color indicator when generation is complete.
ComfyUI utility nodes for preset management including save, load, update and delete operations. (Description by CC)
Inpaint helper nodes for ComfyUI (311 series). First node: Add Margins — pad an image and its mask with extra width/height, centered on a configurable fill color.
Expertly designed custom nodes for Google Gemini integration in ComfyUI, featuring a robust multi-key retry system to eliminate server errors and rate limits.
Structured workflow memory and human-in-the-loop creative feedback for ComfyUI that provides persistent memory across runs with session history, interactive feedback panels, and rating/annotation capabilities.
Publish your ComfyUI outputs to your Fanvue account: upload images and videos to your vault and schedule posts. Needs a free LatentPost account. https://latentpost.com/
ComfyUI nodes for Kinamix concept embeddings
Links the generated image to the references
PIQ ComfyUI custom nodes
LLM nodes for ComfyUI including a Canary-Qwen-2.5B audio transcription node. (Description by CC)
A collection of custom nodes for ComfyUI
Custom node for ComfyUI that performs speaker diarization with chronological output ordering.
All-in-One collection of Painter's ComfyUI nodes for video generation
A dependency-friendly translation node pack for ComfyUI, connecting Google, Baidu, Ollama, OpenAI-compatible models, and custom HTTP APIs through one unified interface.
Custom nodes for sampling RUM Anima x-pred checkpoints in ComfyUI.
ComfyUI nodes to use [a/DeepSeek-VL2](https://github.com/deepseek-ai/DeepSeek-VL2)
The ComfyUI version of [a/XVerse](https://github.com/bytedance/XVerse)
A custom node that crops images to specified dimensions from a center point with optional offset
ComfyUI nodes for ChatGPT and Grok prompt, image, and video workflows via an OpenAI-compatible gateway.
ComfyUI model dependency manifests, selected-model downloads before execution, and ModelScope batch imports.
Apply local prompt skills for MiniMax H3 video prompts and Music 3 music descriptions in ComfyUI. Text generation requires an external API key.
3D previsualisation and spatial layout node for ComfyUI. Beauty, depth, and normal passes from a Three.js viewport.
Nodes for incrementally constructing large, multifaceted images
Power-user FFmpeg toolkit for ComfyUI. Tensor<->media bridge with broadcast-grade codec control: subtitle burn-in, multi-overlay Compose pipeline, provider-agnostic AI transcribe/translate, long-video timeline with latent carry, NVENC-aware. Depth > breadth.
ComfyUI node for Flux Context (Kontext) image editing
ComfyUI nodes that call GPT / Gemini / Grok / Qwen image and text models through a user-supplied API endpoint (official API, relay or private proxy).
ComfyUI node which allows you to save generated images as .ICO files.
ComfyUI node for the fixed BytePlus Seedance 2.0 model through Vergora
ComfyUI node for BytePlus Seedance 2.0 Fast and Mini through Vergora
ComfyUI node for BytePlus Seedance 2.5 video generation through Vergora
Logic, conditional and dynamic loader nodes for ComfyUI.
Multi-session DeepSeek prompt chat for ComfyUI, with TXT knowledge retrieval, bilingual prompt tags, and confirmed English prompt output.
ComfyUI custom nodes for RunningHub cloud workflow execution and AI model invocation
An isolated SM89 safety patch for MiniMax H3 long-sequence SageAttention.
ComfyUI custom node: Real-ESRGAN video upscale with MediaPipe hand region enhancement for real-person dance footage
you can using EchoMimic v1,v2,v3 version in comfyui,please using pip install install miss module
ComfyUI node that computes relative camera movement between two extrinsics matrices and injects it into a prompt template as JSON.
Custom node that migrates WebUI Forge's Ranbooru plugin to ComfyUI, enabling searches across Booru sites to generate positive/negative prompts with tag cleaning, rating sorting, and caching support. (Description by CC)
Nodes: Center Calculation. Improved Numerical Calculation for ComfyUI
Helper for ComfyUI-HairDeck: runs it on RTX 20 series and older (FP32 fallback), without triton-windows and with setuptools 81+. Adds no nodes and does not change HairDeck's files.
Sound effects from an English prompt in ComfyUI with MOSS-SoundEffect v2.0. Runs locally, works on 8 GB cards.
ComfyUI custom node pack for prompt lists: build a list once, edit its slots in place, read items by index, and flatten it back into one prompt line.
A ComfyUI custom node that sends workflow output (images and videos) via Telegram Bot to specific users. (Description by CC)
ComfyUI nodes for DeepSeek API: chat (text + images + web tools), web search, and account balance
Extract PurgeVRAMNode from ComfyUI-tbox for easy VRAM management. (Description by CC)
This custom node returns one uniformly selected integer from an inclusive range.
A conditional branch node for ComfyUI that inputs an INT or STRING and selects output paths by multiple conditions. (Description by CC)
This is a custom node for ComfyUI that calls the DeepSeek Chat API to process text input and return text output.
A comprehensive collection of ComfyUI nodes designed to reduce reliance on multiple third-party node packages.
Deterministic lace scan cleanup, repeat reconstruction, background replacement, and mask output for ComfyUI.
Enables wildcard substitution, multiple anchoring
Schema-first input/output parameter nodes for ComfyUI workflows.
BluePixel-Turbo image editing nodes: 4-step distilled Flux2-Klein + DCAE at a 4K bucket area, with FP32 and INT8 transformer support.
A ComfyUI server plugin that adds server-side model downloading capability to the missing models dialog, enabling backend downloads and automatic model categorization. (Description by CC)
AI video creation wizard — chat with AI director, analyze images, generate previews, and render with Seedance 2.0, Kling, Veo, and 15+ models — all inside ComfyUI
ComfyUI nodes for LivePair AI — private image & video generation API (pay per result, no subscription)
Save ComfyUI images directly into a bundled Artifex gallery.
ComfyUI custom nodes for running prompt sequences from one queue.
Pixaroma-style LoRA stacks with workflow-wide category switches, trigger filtering, and reusable per-LoRA presets. Requires ComfyUI-Pixaroma.
A ComfyUI custom node that sanitizes strings for use as filesystem paths
ComfyUI nodes to use the SeeCoder from [a/Prompt-Free-Diffusion](https://github.com/SHI-Labs/Prompt-Free-Diffusion)
Custom Nodes for ComfyUI to Simplify Multi-Resolution and Prompt Workflows
ApplySpectralLock — latent spectral handoff for Flux.1 / Rectified Flow and USDU tiled upscaling.
Flux self-guidance node that adds micro-detail during a refine without raising denoise or changing composition. Runs inside tiled samplers.
Single-node, 3-stage Flux pipeline: base generation, progressive locked-noise upscale, and an optional refine pass. Driven by any SAMPLER (e.g. RES4LYF ClownSampler) and preserves composition across resolution changes via orthogonal subspace noise locking.
Exposes eta/s_noise/r_1/r_2 controls for ComfyUI's built-in seeds_3 solver. eta=0 gives a clean deterministic 3rd-order exponential sampler that preserves structure/text at high denoise without grain.
ComfyUI custom nodes for neural color match: two images in, graded source out.
Nodes:Save Image with Options
A ComfyUI custom node pack for derp-themed workflow utilities, loaders, dockable UI panels, palettes, and prompt tools.
Lostless custom nodes for random image selection, sequence buffering, mask editing, and image-to-mask conversion in ComfyUI.
Check any workflow for the custom nodes and model files your install is missing. Adds a sidebar panel and a node that diff a workflow against the node types and model files you actually have, with HuggingFace/CivitAI search links for every missing model.
Practical ComfyUI utility nodes for folder I/O, image tools, color matching, seamless textures, prompt conversion, performance tracking, and optional QwenVL screening.
Replace part of an image with another image
ComfyUI nodes for Jina CLIP v2 model loading, text encoding, and SDXL adapter support. (Description by CC)
Additional nodes for use alongside ComfyUI_LLM_SDXL_Adapter with enhanced T5-GEMMA text encoding and improved adapter loading capabilities.
Custom nodes to modify the size and crop conditioning for SDXL.
This is a test nodepack for safely testing the functionality of ComfyUI-Manager. Please do not install this nodepack.
This is a test nodepack for safely testing the functionality of ComfyUI-Manager. Please do not install this nodepack.
The DPRW (Diffusion-based Perceptual Robust Watermarking) node is a custom node designed for **ComfyUI**, aimed at protecting generated images by embedding watermarks into latent noise. It leverages the characteristics of diffusion models to embed and extract watermark information without significantly affecting image quality. The DPRW node provides three main functional categories, assisting users in adding watermarks, extracting watermarks, and performing advanced sampling during the image generation process.
🌐 ARMD (Adaptive Regional Mixture of Diffusers) is a region-based diffusion system for ComfyUI, designed for semantic upscaling with per-region conditioning, optional VLM captioning, and adaptive regional reconstruction.
Image Tile Split with Region-Aware Prompting for ComfyUI
Tiled regional prompting + tiled VAE decode with seam-free blending for ComfyUI
ComfyUI node and CLI for offline image-batch preflight reports.
String utility nodes for ComfyUI: attach titles to prompts, loop them through your workflow, assemble save paths, and derive a distinct seed per image.
INT+ node for ComfyUI with control after generate style functionality for signal-controlled increment/decrement/randomize operations.
A package implementing a Bitalino device ComfyUI custom node.
ComfyUI custom node collection including face direction detection, image processing, and example nodes. (Description by CC)
Experimental ComfyUI custom node wrapper for ByteVisionLab/DreamLite, providing text-to-image and text-guided image editing through official pipeline API.
Model-agnostic workflow controllers for ComfyUI Nodes 2.0: bind arbitrary nodes, use Mute/Bypass, Solo/Restore, and switch exclusive workflow branches.
Tiny ComfyUI helper nodes: ListToAudio + ConcatImageBatches for multi-segment video stitching
ComfyUI layout toolbar with exact node spacing, reference alignment, move and resize snapping, edge snapping, collapsed and pinned node support, and horizontal/vertical fill tools.
Compressed-sensing hybrid 4K image-to-video: RAFT flow transfer, dual-channel Laplacian recompose, CS tile scheduler, keyframe refresh helpers.
Encrypt sensitive node parameters in ComfyUI workflows. PBKDF2 + AES-256-GCM, password never leaves your browser.