I’m quite late to the game, but I decided to try out image generation AI and installed ComfyUI on Ubuntu.

Basically, I just needed to follow the ComfyUI Wiki ⧉.

comfycli

python -m venv comfy-env
source comfy-env/bin/activate
pip install comfy-cli
comfy --install-comp
comfy install
pip3 install torch torchvision torchaudio
comfy launch -- --listen 0.0.0.0

It was installed in ~/comfy/ComfyUI/. By using --listen 0.0.0.0, I can access it from a browser on another PC via ip address:8188.

Auto-start with systemd

Create a startup script

vim ~/comfy/comfyui-start.sh

#!/bin/bash
cd /home/myusername/comfy
source /home/myusername/.bashrc
source /home/myusername/comfy-env/bin/activate
comfy launch -- --listen 0.0.0.0

chmod +x ~/comfy/comfyui-start.sh

Create a systemd unit file

sudo vim /etc/systemd/system/comfyui.service

1
[Unit]
2
Description=ComfyUI Service
3
After=network.target
4

5
[Service]
6
Type=simple
7
User=myusername
8
ExecStart=/home/myusername/comfy/comfyui-start.sh
9
Restart=always
10
WorkingDirectory=/home/myusername/comfy/ComfyUI
11

12
[Install]
13
WantedBy=multi-user.target

Register with systemd

sudo systemctl daemon-reexec
sudo systemctl daemon-reload
sudo systemctl enabel comfyui
sudo systemctl start comfyui
systemctl status comfyui

Installing ComfyUI on Ubuntu 24.04

comfycli

Auto-start with systemd

Create a startup script

Create a systemd unit file

Register with systemd

Author

Fumito Iriya