Build Your Private AI Image Generator: Docker Model Runner + Open WebUI

Introduction

Have you ever needed to generate images for a project but hesitated because of privacy concerns, credit limits, or frustrating content filters? You type in a prompt for a dragon in a business suit, and the service rejects it for being “unsafe.” Meanwhile, you’re left wondering where your data ends up. What if you could eliminate all those worries by running the entire image generation pipeline on your own machine—with a clean, chat-based interface on top?

Build Your Private AI Image Generator: Docker Model Runner + Open WebUI
Source: www.docker.com

That’s exactly what Docker Model Runner now makes possible. With just a few commands, you can download an image generation model, connect it to Open WebUI, and start creating images directly from a chat interface—completely local, fully private, and entirely under your control. Think of it as your own private DALL-E, with no cloud subscription and no data leaving your computer.

What You’ll Need

Before we begin, make sure your environment meets these requirements:

If you can run docker model version without errors, you’re good to go. The process is designed to be straightforward, even if you’re not a machine learning expert.

How Docker Model Runner Works with Open WebUI

Before diving into the steps, let’s understand the architecture. Docker Model Runner acts as the control plane. It downloads the model, manages the inference backend lifecycle, and exposes a 100% OpenAI-compatible API. This includes the /v1/images/generations endpoint that Open WebUI already knows how to communicate with. In other words, the integration is seamless—Open WebUI thinks it’s talking to a cloud service, but everything runs locally.

Step 1: Pull an Image Generation Model

Docker Model Runner uses a compact packaging format called DDUF (Diffusers Unified Format) to distribute image generation models via Docker Hub, just like any other OCI artifact. This single-file format bundles all components of a diffusion model (text encoder, VAE, UNet/DiT, scheduler config) into one portable package that Docker Model Runner can unpack at runtime.

Start by pulling the stable-diffusion model:

docker model pull stable-diffusion

Once the download completes, verify the model is ready:

docker model inspect stable-diffusion

You’ll see output similar to this:

{
    "id": "sha256:5f60862074a4c585126288d08555e5ad9ef65044bf490ff3a64855fc84d06823",
    "tags": ["docker.io/ai/stable-diffusion:latest"],
    "created": 1768470632,
    "config": {
        "format": "diffusers",
        "architecture": "diffusers",
        "size": "6.94GB",
        "diffusers": {
            "dduf_file": "stable-diffusion-xl-base-1.0-FP16.dduf",
            "layout": "dduf"
        }
    }
}

The model is stored locally as a DDUF file. This means you won’t have to manage multiple files or dependencies—everything is in one place, and Docker Model Runner knows how to load it when needed.

Build Your Private AI Image Generator: Docker Model Runner + Open WebUI
Source: www.docker.com

Step 2: Launch Open WebUI

This is where the magic happens. Docker Model Runner includes a built-in launch command that automatically wires up Open WebUI against your local inference endpoint. Just run:

docker model launch openwebui

That’s it. Behind the scenes, Docker Model Runner starts the inference backend (using the model you pulled in Step 1), configures it to listen on the correct port, and launches Open WebUI configured to communicate with that local endpoint. You’ll be greeted with a chat interface, ready to accept your image generation prompts.

No need to set up environment variables, no complicated YAML files—just one command that handles everything. You can now type prompts like “a dragon wearing a business suit” and watch as Stable Diffusion creates the image entirely on your machine.

Conclusion

By combining Docker Model Runner with Open WebUI, you gain complete control over AI image generation. No more worrying about cloud privacy, credit exhaustion, or absurd content filters. Your prompts stay on your computer, and your images are generated locally using powerful open-source models like Stable Diffusion.

This setup is perfect for hobbyists, developers, and anyone who values privacy while exploring creative AI. With just two commands—docker model pull stable-diffusion and docker model launch openwebui—you’re ready to go. So why not give it a try? Your own private image generator is just a few keystrokes away.

Recommended

Discover More

Google's AI Search Now Highlights Reddit and Forum Discussions as 'Expert Advice'Bleeding Llama Vulnerability Threatens 300,000 Ollama AI Deployments — Remote Exploit Without AuthenticationRocsys Unveils Autonomous Charging for Robotaxi Fleets, Secures $13M in New FundingThe Lingering Legacy of Leaded Aviation FuelFintech Product Failures Linked to Feature Overload: Experts Urge Shift to 'Bedrock' Strategy