Why is ComfyUI slow on Mac?

Make sure you selected Apple Metal during setup, which enables GPU acceleration. If you installed via pip without Metal support, reinstall PyTorch using the Apple Metal nightly build.

How to Install ComfyUI on Mac (Apple Silicon)

Two methods covered: (1) The official Desktop app — easiest, recommended for all Apple Silicon Macs. (2) The manual GitHub install — more control, works on Intel Macs too. Jump to whichever fits your situation.

Desktop App Install (Easy — Recommended for Apple Silicon)

The official ComfyUI Desktop app is the fastest way to get started on an Apple Silicon Mac (M1, M2, M3, M4). It handles Python, PyTorch, and all dependencies automatically — no Terminal required.

What you need

A Mac with Apple Silicon (M1, M2, M3, or M4).
The Desktop app does not support Intel Macs — use the manual install instead.

⚠️ Warning: If macOS blocks the app on first launch, go to System Settings → Privacy & Security and click "Open Anyway".

Step-by-step installation

Download the Desktop app

Go to comfy.org/download and download the Mac Desktop (.dmg) file.

ComfyUI download page showing the Mac Desktop DMG option — Download the Mac Desktop (.dmg) from comfy.org/download

Install the app

Open the downloaded .dmg file. Drag ComfyUI to the Applications folder.

ComfyUI DMG window showing drag-to-Applications install

Open ComfyUI

Open ComfyUI from your Applications folder (Spotlight search works too). Click "Get started" on the welcome screen.

ComfyUI welcome screen showing the "Get started" button

Select Apple Metal

Choose the Apple Metal option. This enables GPU acceleration on your Mac. Click Next.

ComfyUI setup screen showing the Apple Metal GPU option selected

Choose install location

Pick where to install ComfyUI. The default location is fine for most users. Click Next.

ComfyUI setup screen showing the install location selection

Optional settings

Enable automatic updates if you want. Usage metrics are optional. Click Next.

ComfyUI setup screen for automatic updates and usage metrics settings

Click Install and wait

Click Install. The app will download and install all required components. This may take a few minutes. The app will open automatically when done.

ComfyUI installation progress bar screen

Start generating images

Choose a template like "image generate" from the template picker and start making images.

ComfyUI template picker showing the image generate workflow option

💡 Tip: If you see a blank canvas after the app opens, click Load Default or select a template from the menu to get a working workflow immediately.

Video walkthrough

Prefer to watch? The full Desktop app installation is shown step by step in this video:

Manual Install via GitHub (More Control)

The manual install gives you full control over Python version, environment, and ComfyUI configuration. Use this method if you want custom setups, are on an Intel Mac, or prefer the command line.

Prerequisites

Homebrew (brew) — Mac package manager
Git — for cloning the repository
Python 3.10–3.12 (this guide uses Python 3.12)
Miniconda or Python venv for environment management (Miniconda gives better Apple Silicon PyTorch support)

Step 1 — Install Homebrew and Git

First, check if Homebrew is already installed:

brew --version

If Homebrew is not installed, run:

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

Then add Homebrew to your PATH (required on Apple Silicon):

eval "$(/opt/homebrew/bin/brew shellenv)"

Terminal showing Homebrew installation output on Mac Apple Silicon — Homebrew installing in Terminal

Check if Git is installed:

git --version

If Git is missing:

brew install git

Terminal showing git installation via brew on Mac — Installing Git via Homebrew

Step 2 — Clone ComfyUI

Navigate to the folder where you want ComfyUI installed, then clone the repository:

git clone https://github.com/Comfy-Org/ComfyUI.git

Terminal showing git clone of ComfyUI repository — Cloning the ComfyUI repository

Enter the project folder:

cd ComfyUI

Step 3 — Install Python and Set Up an Environment

You have two options: Option A uses Python venv and pip (simpler), Option B uses Miniconda (recommended for Apple Silicon PyTorch support). See Apple's PyTorch Metal guide for additional reference.

Option A — Python venv + pip

Install Python 3.12 if not already available:

brew install python@3.12

Create and activate a virtual environment inside the ComfyUI folder:

python3.12 -m venv comfyuienv

Terminal showing Python venv creation for ComfyUI — Creating the Python virtual environment

source comfyuienv/bin/activate

Install PyTorch with Apple Metal support:

pip install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu

Terminal showing PyTorch nightly installation for Apple Silicon Mac — Installing PyTorch with Apple Silicon support

Option B — Miniconda (Recommended for Apple Silicon)

Download and run the Miniconda installer for Apple Silicon:

curl -O https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-arm64.sh

sh Miniconda3-latest-MacOSX-arm64.sh

Follow the installer prompts, then restart your terminal or run source ~/.zshrc.

Create a conda environment and install PyTorch with Metal support:

conda create -n comfy python=3.12

conda activate comfy

conda install pytorch torchvision torchaudio -c pytorch-nightly

Step 4 — Install Requirements and Run

From the ComfyUI folder (with your venv or conda environment active), install all dependencies:

pip install -r requirements.txt

Then start the ComfyUI server:

python main.py

Open your browser and go to: http://127.0.0.1:8188

ComfyUI web interface loaded in a browser at localhost 8188 — ComfyUI running in the browser at http://127.0.0.1:8188

💡 Tip: Add --force-fp16 after python main.py if you encounter memory issues on lower-RAM Apple Silicon Macs.

Common Problems and Fixes

If something goes wrong during the manual install, one of these is almost always the cause:

Port 8188 already in use

Fix: Another application is already using port 8188. Find and stop it, or start ComfyUI on a different port with: python main.py --port 8189

"torch" import errors on startup

Fix: Your PyTorch build does not match your Python version or CPU architecture. Make sure you installed a PyTorch build compiled for arm64 (Apple Silicon) — not the x86 Intel build.

Permission errors when installing packages

Fix: Run your commands inside a folder you own. Avoid /root or system directories. If using pip, make sure your venv or conda environment is active before running pip install.

Strange crashes or unexpected errors

Fix: Delete your venv or conda environment and recreate it from scratch. A clean environment resolves most dependency conflicts. Remove the venv folder and run the setup commands again.

Desktop app blocked by macOS Gatekeeper

Fix: Go to System Settings → Privacy & Security. Scroll down and click 'Open Anyway' next to the ComfyUI entry. This is a standard macOS security prompt for apps not from the App Store.

Troubleshooting Checklist

Run through this checklist before searching for help:

Homebrew missing

Install it and run: eval "$(/opt/homebrew/bin/brew shellenv)"

Git missing

Run: brew install git

Wrong Python version

Install Python 3.12 via Homebrew or use a conda environment

ComfyUI server not showing in browser

Check terminal output from python main.py for startup errors

Desktop app blocked by macOS

Open System Settings → Privacy & Security → click "Open Anyway"

Slow generation / no GPU acceleration

Confirm you selected Apple Metal during setup. Reinstall PyTorch for arm64 if needed.

Missing custom nodes after install

Install ComfyUI Manager to browse and install nodes from within the interface

Need deeper help? ComfyUI Troubleshooting Guide covers a comprehensive list of errors and fixes across all workflow types and operating systems.

Frequently Asked Questions

Yes. ComfyUI works on Mac with Apple Silicon (M1, M2, M3, M4) using either the official Desktop app or a manual GitHub install. Intel Macs are not supported by the Desktop app but can run ComfyUI via the manual install method.

The easiest method is the official ComfyUI Desktop app from comfy.org/download. It handles Python, PyTorch, and all dependencies automatically — no Terminal required. Recommended for all Apple Silicon users.

The Desktop app does not support Intel Macs. You can still run ComfyUI on an Intel Mac using the manual GitHub install with a Python venv or Miniconda environment.

Python 3.10 to 3.12 is recommended. This guide uses Python 3.12 for the best compatibility with current PyTorch builds. Avoid Python 3.13+ for now as some dependencies are not yet compatible.

Make sure you selected Apple Metal during Desktop app setup, or installed the arm64 PyTorch nightly build for the manual install. Without Metal, generation runs on CPU only, which is much slower. The --force-fp16 flag can also help on lower-RAM Macs.