Here's how you can easily run ComfyUI-Launcher on Runpod.
- Create a new pod with this template: https://runpod.io/console/gpu-cloud?template=lr4wb7ypuy&ref=pc9tg9pr
- If you want to save your data across pods, mount a Network volume at
/workspace
. - Pick a GPU & region, and Deploy the pod.
Once the pod is running, click the Connect button and click the first Connect to HTTP Service [Port 4000] button.
This will open a new tab with ComfyUI-Launcher running.
When creating/importing workflow projects, ensure that you set static ports, and ensure that the port range is between 4001-4009
(inclusive). This is required for ComfyUI-Launcher to work properly with Runpod.
Your models + project files will be saved under /workspace/comfyui_launcher_models
and /workspace/comfyui_launcher_projects
, respectively.
SSH into the Runpod instance to view & manage these folders.