- Home
- Install
- Install Comfyui
- Install ComfyUI On Mac
Install ComfyUI On Mac
Step-by-step guide to install ComfyUI on macOS. Covers both Portable and Desktop versions, Homebrew setup, Python dependencies, and troubleshooting tips for Apple Silicon and Intel Macs.
System Requirements for Mac
| Component | Requirement |
|---|---|
| System | macOS 13+ (Ventura) |
| Chip | Apple Silicon (M1/M2/M3/M4) recommended; Intel Macs supported with limited performance |
| RAM | 8 GB minimum, 16 GB recommended |
| Storage | SSD with at least 20 GB free space (bare minimum for ComfyUI itself — actual model downloads can reach 50–100GB+ quickly) |
| Browser | Chrome 143+ recommended |
If you are using a Mac, the best way to run ComfyUI depends on whether your Mac has an Apple Silicon chip or an Intel chip.
ℹ️ Note about updates: For the most up-to-date Desktop installation info, check the official docs.
Comfy Desktop (Recommended)
Comfy Desktop is the official desktop application and the easiest way to run ComfyUI on macOS. It handles the installation, Python environment, and dependencies automatically with Metal Performance Shaders (MPS) for GPU acceleration on Apple Silicon.
→ Comfy Desktop Installation Guide
Comfy CLI
You can also use Comfy CLI to install ComfyUI on your Mac. It works on both Apple Silicon and Intel Macs with Python 3.12 or 3.13.
Manual Installation
For advanced users who prefer full control, clone the repository from GitHub and set up the environment manually:
# 1. Clone the repository
git clone https://github.com/comfyanonymous/ComfyUI.git
cd ComfyUI
# 2. Create a virtual environment
python3 -m venv .venv
# 3. Activate the environment
source .venv/bin/activate
# 4. Install PyTorch with Metal support (Apple Silicon)
pip install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu
# 5. Install ComfyUI dependencies
pip install -r requirements.txt
# 6. Start ComfyUI
python main.pyOn Apple Silicon Macs, ComfyUI uses Metal Performance Shaders (MPS) for GPU acceleration, which provides good performance for most models.