How to Run ComfyUI

Learn different ways to start ComfyUI.

Method 1: Direct Run

Windows

# Navigate to ComfyUI folder
cd path/to/ComfyUI

# Run with Python
python main.py

# Or run with python3
python3 main.py

Linux/Mac

cd path/to/ComfyUI
python3 main.py

Method 2: With Custom Port

# Specify port
python main.py --port 8189

# Access at http://127.0.0.1:8189

Method 3: Listen on All Interfaces

# Allow external access
python main.py --listen 0.0.0.0

# Access from other devices on network

Method 4: Enable Auto-launch Browser

python main.py --auto-launch

Command Line Options

Option	Description	Example
--port	Set port number	--port 8189
--listen	Listen on all interfaces	--listen 0.0.0.0
--auto-launch	Auto-open browser	--auto-launch
--force-fp16	Force FP16 precision	--force-fp16

Using Virtual Environment

Create venv

# Windows
python -m venv venv
venv\Scripts\activate

# Linux/Mac
python3 -m venv venv
source venv/bin/activate

Install Dependencies

pip install -r requirements.txt

Run

python main.py

Verification

Check if ComfyUI is running:

1. Open browser
2. Visit http://127.0.0.1:8188
3. You should see the ComfyUI interface

Troubleshooting

Port Already in Use

# Use different port
python main.py --port 8190

Dependencies Missing

# Reinstall dependencies
pip install -r requirements.txt --force-reinstall

GPU Not Detected

Check CUDA installation:

python -c "import torch; print(torch.cuda.is_available())"

Next Steps

• First Image - Generate your first image
• Share Models - Share models between systems