How to install ComfyUI on Windows (2026): step by step

How to install ComfyUI on Windows (2026): step by step

Updated September 2026 · Instructions tested on ComfyUI v0.37. The download links always point to the latest official release. Part of our AI and ComfyUI tutorials.

Installing ComfyUI for the first time can feel a bit intimidating: folders, different versions, words like “portable” and “CUDA”. We've been there too, and today it's the program we use to create most of our AI images and videos, on an ordinary 12 GB RTX 3060.

The good news is that in 2026 it's much easier than it used to be, and you can have your first image in half an hour. In this guide we'll walk you through it step by step: which version to download for your graphics card, how to launch it, how to enable the Manager to install nodes and models, and how to generate your first image. At the end you'll find 3 free, ready-to-use workflows we put together ourselves.

Quick summary

  1. Download the Desktop app (the easiest option) or the portable version that matches your graphics card: all the official links are in the table below.
  2. Launch ComfyUI: it opens in your browser at http://127.0.0.1:8188.
  3. Download a model or load one of our workflows and generate your first image.

Video tutorial: installing ComfyUI on Windows (video in Italian)

The video is in Italian and shows an earlier version of ComfyUI: some screens may look different, but the main steps are the same. You'll find the up-to-date instructions in the text below.

What is ComfyUI

ComfyUI is a free, open-source program for running AI models that generate images, video and audio on your own computer: Stable Diffusion, SDXL, Flux and many newer models. It's node-based, like Blender or Unreal Engine: each step of the generation is a block you connect to the others, and the whole set of blocks is called a workflow.

The big advantage is that almost every new AI model is made available as ComfyUI nodes: instead of installing a different program for each model, ComfyUI is all you need. And workflows can be saved, shared and reused with a single click.

ComfyUI interface with a node-based workflow

ComfyUI download: which version to choose

VersionWho it's forOfficial download
Desktop appBeginners: it installs like any normal program and updates itself. Windows and Mac, NVIDIA, AMD, Intel and Apple Siliconcomfy.org/download
Portable NVIDIANVIDIA RTX 20, 30, 40 and 50 series cardsComfyUI_windows_portable_nvidia.7z
Portable NVIDIA (older cards)Only NVIDIA GTX 10 series and older cardsComfyUI_windows_portable_nvidia_cu126.7z
Portable AMDAMD Radeon cardsComfyUI_windows_portable_amd.7z
Portable IntelIntel Arc cardsComfyUI_windows_portable_intel.7z

The portable links are the project's official GitHub links and always download the latest version (about 2 GB). Heads up: the package for older NVIDIA cards should not be used on the 20 series and newer.

Requirements

  • Windows 10 or 11 64-bit, with up-to-date graphics drivers.
  • Graphics card: an NVIDIA RTX with at least 8 GB of VRAM is recommended; Flux and video models need 12–16 GB or more. It also works with AMD and Intel cards, and even on the CPU alone, but much more slowly.
  • RAM: at least 16 GB, 32 GB is better.
  • Disk space: about 5 GB for ComfyUI, plus room for the models, which take anywhere from 2 to over 20 GB each. An SSD is best.

If you're thinking about buying a graphics card for AI, our guide to the best GPUs for AI covers the recommended models and how much VRAM you need (in Italian).

Method 1: install the ComfyUI Desktop app

  1. Go to comfy.org/download and download the Windows installer.
  2. Run the installer and follow the setup wizard: it asks which graphics card you use and which folder to store models and files in.
  3. When it's done, ComfyUI opens in its own window, ready to go and with the Manager already included. Updates arrive automatically.

This is the easiest route. The portable version is still the favorite of anyone who wants more control, several separate installs, or to run ComfyUI from an external drive.

Method 2: install ComfyUI portable

The portable version already contains everything it needs, Python included: you don't need to install Python, Git or anything else, it doesn't change any system settings, and you can move it to another drive or computer.

  1. Download the package for your graphics card from the download table.
  2. Extract the .7z file to any folder you like, ideally on an SSD with plenty of free space (models take up a lot). On Windows 11 you can extract it straight from File Explorer; otherwise use 7-Zip. If extraction fails: right-click the file › Properties › tick Unblock.
  3. Open the ComfyUI_windows_portable folder and double-click run_nvidia_gpu.bat (AMD and Intel cards have their own matching launch file; run_cpu.bat uses the CPU only).
  4. After a few seconds ComfyUI opens in your browser at http://127.0.0.1:8188. Keep the black command prompt window open: if you close it, ComfyUI closes too.

To update the portable version, use the update\update_comfyui.bat script in the same folder.

Enable ComfyUI Manager

The Manager lets you install custom nodes and models with one click, update everything, and automatically install missing nodes when you open a workflow downloaded from the internet. It's now built into ComfyUI: in the Desktop app it's already on, while in the portable version you only need to enable it once.

  1. Close ComfyUI. In the ComfyUI_windows_portable folder, type cmd in the File Explorer address bar and press Enter: a command prompt opens already in the right folder.
  2. Install the Manager's components with this command:
    .\python_embeded\python.exe -m pip install -r ComfyUI\manager_requirements.txt
  3. Open run_nvidia_gpu.bat in Notepad and add --enable-manager to the end of the first line, like this:
    .\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build --enable-manager
  4. Save and launch ComfyUI: the Manager button appears in the top bar.

Generate your first image

ComfyUI doesn't come with any models, so the first step is to download one. The quickest way is to start from the built-in templates (Workflow menu), which offer to download the required models automatically. Alternatively, download a model and put it in the right folder: Stable Diffusion 1.5 is perfect to start with, light and fast even on cards with little VRAM; copy it into ComfyUI\models\checkpoints.

Then type a description of the image in the prompt node and press Run (or Queue). Every image you generate is saved in the ComfyUI\output folder.

🔥 Download 3 free ready-to-use workflows

To get you started right away we've prepared 3 workflows with built-in notes, made only with ComfyUI's core nodes (so they don't stop working when some third-party node gets abandoned):

  • Stable Diffusion 1.5, standard and LCM for fast generation;
  • SDXL, standard and Lightning for fast generation;
  • Flux Schnell, which generates images in 4 steps.

They're designed to run even on PCs with little VRAM (4–8 GB): on a 12 GB RTX 3060 an image takes just a few seconds. Enter your email below and you'll get the download link straight away; then drag the workflow file into the ComfyUI window and follow the notes (the notes are written in Italian).

Workflow x Comfyui – Stable Diffusion 1.5 + SDXL + Flux Schnell + upscale

Cosa troverai in questo download:- 3 workflow per comfyui per la generazione di immagini con i modelli:
  • 1.5 (standard e LCM per la generazione veloce)
  • SDXL (standard e Lightening per la generazione veloce)
  • Flux Schnell (generazione in 4 step)
- 1 cartella con lo script per installare Comfyui Manager, fondamentale per gestire Comfyui

Common problems and fixes

ProblemFix
ComfyUI won't start or closes immediatelyUpdate your graphics drivers: the NVIDIA portable version needs recent drivers.
c10.dll errorInstall the Microsoft Visual C++ Redistributable package.
Red error in the interface on first launchA model is missing: download one and copy it into ComfyUI\models\checkpoints.
Red nodes when you open a downloaded workflowSome custom nodes are missing: open the Manager, use the option to install missing nodes, then restart.
Memory error (out of memory)Your graphics card doesn't have enough VRAM for that model: use a lighter version of the model (fp8 or GGUF), lower the resolution or close other programs.
You have an older NVIDIA GTX 10 series cardUse the "cu126" portable package listed in the download table.

Choosing a graphics card for ComfyUI

For ComfyUI what matters most is the card's VRAM: with a 12 GB RTX 3060, the cheapest card we recommend and the one we built our workflows on, you can run SDXL and lighter versions of Flux without trouble; with 16 GB you can comfortably work with Flux and video models too. These are the cards we recommend, from the most affordable to the most powerful.

  • NVIDIA RTX 3060 12 GB – the budget choice we use every day
  • NVIDIA RTX 5060 Ti 16 GB – the sweet spot: 16 GB for Flux and video models
  • NVIDIA RTX 5070 Ti 16 GB – much faster, for those who generate a lot
  • NVIDIA RTX 5090 32 GB – top of the line, for the largest models and video
  • Ready-made PC with RTX 5060 Ti 16 GB – if you'd rather not build one yourself

Frequently asked questions

Is ComfyUI free?

Yes, it's free and open source. Most models are free to download too; some have licenses that restrict commercial use, so always check the license of the model you're using.

Desktop app or portable version: which is better?

If you're just starting out, the Desktop app: it installs and updates itself. The portable version is ideal if you want more control, several separate installs, or to keep ComfyUI on an external drive.

Does ComfyUI work without an NVIDIA card?

Yes: there are versions for AMD and Intel cards, and the Desktop app also supports Macs with Apple chips. Without a dedicated graphics card it even runs on the CPU alone, but an image can take minutes instead of seconds.

How much VRAM does ComfyUI need?

4–8 GB for Stable Diffusion 1.5 and SDXL, 12–16 GB to use Flux comfortably, 16 GB or more for video generation models.

How do I update ComfyUI?

The Desktop app updates itself. The portable version is updated with update\update_comfyui.bat, or from the Manager.

Best GPU for AI (2026): ComfyUI, Stable Diffusion and Flux

Best GPU for AI (2026): ComfyUI, Stable Diffusion and Flux

Updated September 2026 · New recommended cards (RTX 50 series) and an updated table showing how much video memory today’s models need. Part of our AI and ComfyUI tutorials (in Italian).

When we started running AI on our own computer, our first question was the same one you keep asking us: “is my graphics card good enough?”. Today we work with ComfyUI on a perfectly ordinary RTX 3060 12 GB, and we can tell you something you won’t find on any spec sheet: video memory matters more than raw power.

VRAM decides which models you can run and at what resolution, and it’s the reason a card from a few years ago with 12 GB can be more useful than a brand-new one with 8. In this guide we explain how much video memory you need, why NVIDIA is the smart choice today and which cards we recommend in 2026, from the cheapest to the most powerful, including the 3060 12 GB we use every day.

In short: the graphics cards we recommend for AI

  • To spend as little as possible: RTX 3060 12 GB, back on sale: the card we run all our tests on, perfect for SDXL and for the lighter versions of Flux.
  • The best choice to get started: RTX 5060 Ti with 16 GB. Be careful not to mix it up with the 8 GB version, which is too limited for AI.
  • For people who use AI every day: RTX 5070 Ti 16 GB, much faster with the same amount of memory.
  • No compromises: RTX 5090 with 32 GB, for the largest video models and for training your own LoRAs.
  • If you need to buy the whole computer: a ready-made PC with an RTX 5060 Ti 16 GB and 32 GB of RAM.

How much video memory do you need for AI?

For running AI locally, video memory (VRAM) matters more than power: the model has to fit into the card’s memory. If it doesn’t, generation slows to a crawl or stops with an out of memory error. A slower card with more memory is often a better choice than a faster one with little memory.

Video memoryWhat you can doVerdict
8 GBStable Diffusion 1.5 and SDXL; Flux only in reduced versions and with patience; video models are almost out of reachJust for trying things out
12 GBSDXL with no trouble, Flux in reduced versions (fp8 or GGUF), short videos with the lightest modelsThe recommended minimum
16 GBFlux and recent image models comfortably, video with models like Wan and LTX-Video in optimized versionsThe ideal choice
24–32 GBThe largest video models at full quality, high resolutions, LoRA trainingFor professionals

Besides video memory, your system RAM matters too: with 32 GB, ComfyUI can offload part of the model from the graphics card to RAM when needed; with 16 GB you hit the limit quickly.

Why an NVIDIA card

Almost all generative AI software and models are developed and optimized first and foremost for NVIDIA cards, thanks to the CUDA platform and Tensor Cores, compute units designed specifically for neural networks. The result is maximum compatibility: when a new model comes out, it works on NVIDIA right away.

ComfyUI now also has packages for AMD and Intel cards, and the Desktop app runs on Macs with Apple silicon: you can use them, but with less compatibility and longer generation times. To give you an idea of the difference, here are the times we measured with ComfyUI on an RTX 3060 12 GB, compared with computers without an NVIDIA card:

Image generated with ComfyUI and SDXL Lightning on an RTX 3060

IMAGE GENERATION TIMES:

ComfyUI – DreamShaper XL model (SDXL Lightning)

  • 3 seconds on an RTX 3060 12 GB
  • 1 and a half minutes on other GPUs (AMD, Intel, M4)

VIDEO GENERATION TIMES:

ComfyUI – LTX-Video 0.9.1

  • 1 minute 20 seconds on an RTX 3060 12 GB
  • over 12 minutes on other GPUs (AMD, Intel, M4)

It’s no coincidence that, as we showed in our infographic on NVIDIA cards with 24 GB of memory (in Italian, based on Italian users), cards with lots of video memory are still uncommon: for most people, 16 GB is the sweet spot.

The best graphics cards for AI in 2026

1. NVIDIA RTX 3060 12 GB: the cheapest option

This is the card we used for our tests, our videos and our free ComfyUI workflows, and it’s back on sale. It’s the least expensive way to get 12 GB of video memory: SDXL runs with no trouble, Flux works in reduced versions (fp8 or GGUF), and so do short videos with lightweight models. It isn’t as fast as newer cards, but for learning and non-professional use it’s an excellent starting point. If your budget allows, the 16 GB of the RTX 5060 Ti will give you more headroom for the future.

2. NVIDIA RTX 5060 Ti 16 GB: the best choice to get started

It’s the cheapest way to get 16 GB of video memory on a new card. It’s not the fastest, but it runs Flux, SDXL and optimized video models without the memory problems of 8 and 12 GB cards, and it uses little power. Make sure you pick the 16 GB version: there’s also an 8 GB RTX 5060 Ti that we don’t recommend for AI.

3. NVIDIA RTX 5070 Ti 16 GB: the perfect balance

The same amount of memory, but far more compute power: images arrive in a fraction of the time, and video becomes genuinely usable for everyday work. This is the card we recommend to anyone who uses AI professionally but doesn’t want to pay flagship prices.

4. NVIDIA RTX 5090 32 GB: the very best

With 32 GB of memory it runs even the largest video models at full quality, generates at high resolutions and lets you train your own LoRAs in reasonable time. It’s expensive and needs a powerful power supply and a roomy case: it makes sense if AI is your job.

How to choose a graphics card for AI

  1. Video memory first: 16 GB if you can, 12 GB at the very least.
  2. Then speed: with the same memory, a higher-tier card generates faster.
  3. Power supply and space: check the manufacturer’s recommended wattage and the card’s length against your case.
  4. The rest of the PC: at least 32 GB of RAM and a large SSD, because models take up tens of GB.

If you want to compare the performance of different cards, you’ll find detailed benchmarks in this Tom’s Hardware article. If instead you need to buy a whole computer, read our guide to the best PC for 3D and rendering, which applies to AI as well.

Alternatives: cloud and online platforms

If you don’t want to buy a new graphics card, you can use AI online. Generation platforms are the simplest option and often give you free credits to start: you’ll find the main ones in our list of AI image generators. Alternatively, you can rent a graphics card in the cloud and run ComfyUI on it, paying by the hour: it’s worth it for occasional use, but it takes a bit more technical know-how.

Got your graphics card? Start here

Frequently asked questions

What is the best graphics card for ComfyUI?

To spend as little as possible, the RTX 3060 12 GB; for most people, an NVIDIA card with 16 GB of memory: the RTX 5060 Ti 16 GB as the balanced choice, the RTX 5070 Ti if you want more speed. The RTX 5090 32 GB is for those working with the largest video models.

Is 8 GB of VRAM enough for Stable Diffusion?

For Stable Diffusion 1.5 and SDXL, yes, but you’re at the limit: newer models and video models require reduced versions and a lot of patience. If you’re buying today, get at least 12 GB, ideally 16.

Can you use AI with an AMD card?

Yes: ComfyUI has a package dedicated to AMD cards. However, compatibility with new models and performance are generally lower than on NVIDIA, and if something goes wrong you’ll find less help in the community.

Is it worth buying a used card?

It can be, to get more memory for the same money, for example an RTX 3060 12 GB or an RTX 3090 24 GB. Only buy from reliable sellers who offer a warranty, and check that the card hasn’t been used for mining.