ComfyUI-Intel-XPU

ComfyUI container with native PyTorch XPU support for Intel Arc GPUs.

  • AI
  • Other
ComfyUI container with native PyTorch XPU support for Intel Arc GPUs. Designed and tested with Intel Arc Pro Battlemage GPUs using the Linux xe driver and Intel Level Zero runtime, it bundles ComfyUI, ComfyUI Manager, the Hugging Face CLI, and aria2, and creates the standard model directories automatically. It lets you run ComfyUI on Intel Arc GPUs in Docker or Unraid without requiring GPU passthrough to a virtual machine.

Platform availability

Available on 1 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

40/100 Fair

Maintained
73last commit 73 days ago
Popular
102 GitHub stars
Easy to install
351 install platform, install notes, Docker image
Light to run
1016 GB minimum RAM

Availability: platform listings confirmed today, links checked 28 days ago

GitHub stars
2
Open issues
2
Last commit
2026-09-09
License
MIT
Activity
stale

Checked 9 days ago - source: GitHub (heroeswearkapes/comfyui-intel-xpu)

Resources & Compatibility

Minimum requirements: 16 GB RAM. ARM (e.g. Raspberry Pi) is not supported.

Checked 54 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Copy the example .env settings file Before starting, run cp example.env .env to create your own .env file, which is the list of settings ComfyUI reads. Without it, you may not set model download tokens or other options, so downloads or features can fail. source
  • Open port 8188 for the web UI In the compose file (the app's startup recipe), keep ports set to 8188:8188 so your browser can reach the ComfyUI control panel. If that port is not opened, the app may run but you won't be able to open it. source
  • Save models, outputs, and settings Map the user, output, and models folders to real folders on your machine so workflows, generated images, and downloaded models survive restarts. If you skip these folder mappings, Docker can delete that data when the container is removed. source
  • Intel GPU must be exposed on the host The Intel GPU must already be working on the Unraid host and exposed through /dev/dri. The exact render device may vary between systems. Verify your Intel GPU render device before installation. source

Fetched 39 days ago - each note links to its source

Alternatives (6)

  • A1111 Stable Diffusion WEB UI - AUTOMATIC1111's Stable Diffusion web UI provides a powerful, web interface for Stable Diffusion featuring a one-click installer, advanced i…
  • Comfy - ComfyUI is an open-source, node-based workflow engine that empowers creators to design, customize, and run complex generative AI pipelines…
  • Draw Things - Draw Things provides a comprehensive but still easy-to-use, mobile and desktop solution for AI-based art generation.
  • InvokeAI - InvokeAI is an open-source, self-hosted Stable Diffusion toolkit for generating and editing images from text or existing images.
  • chaiNNer - A flowchart/node-based image processing GUI aimed at making chaining image processing tasks (especially upscaling done by neural networks)…
  • stability matrix - Multi-platform package manager for Stable Diffusion.

Checked 40 days ago - source: AlternativeTo

Common questions

Does the container provide the Intel GPU driver?

No, the Intel GPU must already be working on the Docker host. The host must have the Linux xe driver and expose the /dev/dri render device.

How do I find the correct Intel render device to expose?

List the devices under /dev/dri and compare their PCI addresses with lspci and udevadm output. Do not assume the device will be renderD131, because numbering varies by system.

Can I customize ComfyUI startup arguments and Python packages?

Yes, optional CLI_ARGS can be used for ComfyUI startup arguments. Optional PIP_PACKAGES can be used to install runtime Python packages.

How are image versions labeled?

Images are published as latest or as versioned tags using the build date and upstream ComfyUI Git commit. An example format is YYYY.MM.DD- followed by the commit hash.

What ComfyUI data is kept persistent?

The image supports persistent model storage, custom nodes, and input/output directories. It also automatically creates standard ComfyUI model directories.

Answers sourced from github.com

Community

Collected 48 days ago - public community threads