Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Yes—you can run a private, browser-based LLM workstation without a paid AI API subscription by combining Open WebUI (the interface) with Ollama or another compatible model server (the model engine). Docker packages the interface, while a persistent volume keeps accounts, chats and settings when the container is replaced.
The standard setup takes one Docker command: generate a secret, map port 3000, mount open-webui:/app/backend/data, and let the container reach host Ollama through host.docker.internal:host-gateway. Your exact command changes if Ollama runs in another container, on another server, through Docker Model Runner, or on an NVIDIA GPU.
As an Amazon Associate I earn from qualifying purchases.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsWhat are Open WebUI and Ollama, and why do you need both?
Open WebUI is the browser application where you create users, start chats, choose models and change settings. It does not provide the LLM weights by itself. Ollama (or an OpenAI-compatible local endpoint) downloads and serves those models, and Open WebUI sends requests to that service.
This separation is useful: you can replace the model server without replacing the interface, or move the model server to a more powerful machine while keeping the same web application. It also means that a running Open WebUI container does not prove that a model is available; the endpoint must be running and reachable.
Popular on Amazon
- Read Before You Buy — No Video Output: These adapters support charging and USB 2.0 data transfer, but cannot transmit video signals. Except for standard USB webcams (which use USB data only), they are not compatible with HDMI/DisplayPort cables, video-capable USB-C hubs, or docking stations with video output.
- Convert USB-A Ports to USB-C: Designed to connect USB-C earphones, cables, flash drives, card readers, and other USB-C accessories to standard USB-A ports. Plug-and-play with no drivers or software required.
- Aluminum Alloy Housing: Built with a sturdy aluminum alloy shell that aids in heat dissipation and protects against daily wear and scratches. Designed to maintain a stable and secure connection.
- Compact & Travel-Friendly: The ultra-compact design allows the adapter to stay plugged into your device without blocking adjacent ports or adding bulk, reducing wear and tear on your original USB ports.
- 12-Month Warranty: Backed by a 12-month manufacturer warranty for peace of mind. Designed to meet strict quality control standards for reliable everyday performance.
The project documentation says Open WebUI works on macOS, Linux (x86_64 and ARM64, including Raspberry Pi and NVIDIA DGX Spark), and Windows. Docker itself must be installed on the machine where you launch the container, and you need an Ollama service on the host, in another container, on another server, or through Docker Model Runner.
What should you prepare before installing?
- Docker Engine or Docker Desktop: confirm that the Docker CLI and daemon are working with
docker version. - An endpoint: install and start Ollama on the host, prepare an Ollama container, provide a reachable remote Ollama URL, or enable Docker Model Runner.
- Persistent storage: Docker will create the named volume used for Open WebUI data. Ollama also needs persistent storage for downloaded model files.
- A secret: generate a random value and keep it private. The secret is used to preserve authentication data when the container is recreated.
- Optional NVIDIA support: for CUDA acceleration, install a compatible NVIDIA driver, the NVIDIA Container Toolkit and Docker GPU support before adding
--gpus all.
Generate a secret in a terminal with:
openssl rand -hex 32Copy the resulting value somewhere secure. Do not publish it in a shell history, compose file committed to a repository, screenshot or support post.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11How do you install Open WebUI with the official Docker command?
For a host-based Ollama service and a normal CPU-capable Open WebUI container, run the following command. Replace your-secret-key with the value generated above.
docker run -d -p 3000:8080
--add-host=host.docker.internal:host-gateway
-v open-webui:/app/backend/data
-e WEBUI_SECRET_KEY=your-secret-key
--name open-webui --restart always
ghcr.io/open-webui/open-webui:main- Open the interface: visit
http://localhost:3000in a browser. If Docker runs on another machine, use that machine’s address and ensure the port is allowed by its firewall. - Create the first account: the first registration becomes the initial account for that Open WebUI instance.
- Check the connection: confirm that the Ollama endpoint is detected and that model information can be retrieved.
- Pull or select a model: use the interface to choose an available model, or manage models through the Ollama service. There is no universal first-run model or guaranteed performance result; model size and hardware determine what is practical.
The 3000:8080 mapping exposes the container’s port 8080 on port 3000 of the Docker host. The open-webui named volume is mounted at /app/backend/data; it is what protects users, chats and settings from ordinary container replacement. The host-gateway entry gives a Linux container a standard name for reaching services on the host.
Keep WEBUI_SECRET_KEY unchanged for the lifetime of the installation. Open WebUI’s documentation warns: “Without a fixed WEBUI_SECRET_KEY, you’ll be logged out every time your container is recreated.”
How do you run the same setup with Docker Compose?
Compose is easier to maintain when you expect to update, recreate or move the service. Save this as compose.yaml:
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Also popular on Amazon
- 5-in-1 USB-C Hub: Experience comprehensive connectivity featuring a Power Delivery input, two USB-A 2.0 ports, a USB-A 3.0 port, and an HDMI port. (Note: The USB-C power delivery input port is only for connecting an external wall charger to power your laptop and cannot power peripheral devices.)
- 90W Pass-Through Charging: Achieve optimal charging with 90W pass-through power to your laptop, supported by a total input of 100W, with the hub reserving 10W for operational efficiency. (Note: Wall charger not included.)
- Quick Data Transfers: Accelerate your productivity with rapid data transfers using a high-speed 5Gbps USB 3.0 port and two 480Mbps USB 2.0 ports.
- 4K HDMI Display: Enhance your visual experience with a hub capable of delivering 4K resolution at 30Hz in both mirror and extend modes. Please note that this hub is compatible with MacBook (macOS 12 and newer), Windows 10 and 11, ChromeOS, and laptops equipped with DP Alt Mode and Power Delivery. Note: This device is not compatible with Linux.
- What You Get: Anker USB-C Hub (5-in-1, 4K HDMI), welcome guide, 18-month warranty, and our friendly customer service.
services:
open-webui:
image: ghcr.io/open-webui/open-webui:main
ports:
- "3000:8080"
volumes:
- open-webui:/app/backend/data
extra_hosts:
- host.docker.internal:host-gateway
environment:
- WEBUI_SECRET_KEY=your-secret-key
restart: unless-stopped
volumes:
open-webui:Replace the placeholder secret before starting it, then run:
docker compose up -dUse docker compose ps to check the container and docker compose logs -f open-webui to watch startup messages. The service remains available at http://localhost:3000.
To update this floating :main image, pull the new image and recreate the service:
docker compose pull
docker compose up -dAdd container_name: open-webui only when another tool specifically expects that exact container name; Compose networking and service discovery do not require it.
Do not use docker compose down -v casually. The -v option removes the named volume and therefore deletes Open WebUI chats, users and settings.
How do you connect Open WebUI to Ollama in each deployment?
The correct endpoint depends on where Ollama is running. Open WebUI can use the default host route, a Docker-network address, a remote URL, or Docker Model Runner.
Also popular on Amazon
- Sleek 7-in-1 USB-C Hub: Features an HDMI port, two USB-A 3.0 ports, and a USB-C data port, each providing 5Gbps transfer speeds. It also includes a USB-C PD input port for charging up to 100W and dual SD and TF card slots, all in a compact design.
- Flawless 4K@60Hz Video with HDMI: Delivers exceptional clarity and smoothness with its 4K@60Hz HDMI port, making it ideal for high-definition presentations and entertainment. (Note: Only the HDMI port supports video projection; the USB-C port is for data transfer only.)
- Double Up on Efficiency: The two USB-A 3.0 ports and a USB-C port support a fast 5Gbps data rate, significantly boosting your transfer speeds and improving productivity.
- Fast and Reliable 85W Charging: Offers high-capacity, speedy charging for laptops up to 85W, so you spend less time tethered to an outlet and more time being productive.
- What You Get: Anker USB-C Hub (7-in-1), welcome guide, 18-month warranty, and our friendly customer service.
Ollama running directly on the host
With the command or Compose file above, leave OLLAMA_BASE_URL unset. The host-gateway mapping allows Open WebUI to look for Ollama at http://host.docker.internal:11434. Make sure Ollama is started on the host and that its API is listening on the expected port.
Ollama in a separate container
Give Ollama its own persistent volume so downloaded model files survive container replacement. A representative NVIDIA-enabled command is:
docker run -d --gpus=all
-v ollama:/root/.ollama -p 11434:11434
--name ollama ollama/ollamaFor CPU-only Ollama, omit --gpus=all. Put both containers on a shared Docker network and use the Ollama service or container name as the reachable host, or set OLLAMA_BASE_URL to the URL that the Open WebUI container can actually reach. The ollama:/root/.ollama volume contains the model files; deleting it forces those downloads again.
Ollama on another server
Set the variable when creating Open WebUI, using the correct internal or public address:
-e OLLAMA_BASE_URL=https://your-ollama-hostFor Compose, add OLLAMA_BASE_URL=https://your-ollama-host under environment. Protect a remote endpoint with network controls and authentication appropriate to your environment. The setup variable only tells Open WebUI where to send requests; it does not create a security boundary or add authentication by itself.
Docker Model Runner
Docker’s documented integration uses the host-gateway route and port 12434:
Free tools Windows power users keep installed
One-click scans. No signup required.
Also popular on Amazon
- Dual Converters, Infinite Potential：Includes 2× USB C male to USB A female adapters and 2× USB A male to USB C female adapters. Perfect for a wide range of uses—tablets with Bluetooth keyboards, expand USB ports on macbook, and more. Two different converters for all your daily needs
- Next-Level 10Gbps & 3A Charging: No more slow 480Mbps, this usb to usb c adapter has a transfer speed of up to 10Gbps, allowing you to do more transferring in less time. This usb adapter fits both USB A and USB C charger, supporting up to 3A fast charging
- Upgraded Exquisite Craftsmanship: With an aluminum alloy housing and metal connector, the usbc to usb adapter is extremely durable and sturdy. Rigorously tested to withstand more than 10,000 times of plugging and unplugging, ensuring long-lasting performance
- Broad Compatible: The usb c to usb adapter widely supports all USB C/ USB A devices like laptops, tablets, cellphones, car chargers, and phone chargers. Such as compatible with MacBook Pro/Air 2023/2022, Thunderbolt 4/3 Devices,Apple MagSafe Watch 9/8/7/SE/Ultra, iPad Pro 2022/2021, Samsung Galaxy S23/S20/S10, and iPhone 17/16/15 Pro. Plug and play
- Please Note: To reach 10Gbps speed, keep the cable under 3.3 ft. For USB A Male to USB C adapters, try flipping the USB C connector. USB C Male to USB A adapters support bidirectional 10Gbps transfer within 3.3 ft
extra_hosts:
- host.docker.internal:host-gateway
environment:
- OLLAMA_BASE_URL=http://host.docker.internal:12434If you use the runner’s OpenAI-compatible API instead, configure:
OPENAI_API_BASE_URL=http://host.docker.internal:12434/engines/v1
OPENAI_API_KEY=your-local-runner-keyUse the key value accepted by your local runner configuration. Do not assume that the Ollama API path and the OpenAI-compatible path are interchangeable.
Which Open WebUI image should you choose?
The image determines bundled functionality and GPU behavior; it does not determine which Ollama models you can download. Choose the smallest image that still includes the features you need.
| Need | Image or configuration | Trade-off |
|---|---|---|
| Standard installation with bundled helper models | ghcr.io/open-webui/open-webui:main |
Largest standard image; includes bundled speech-to-text and embedding models. |
| Smaller download and memory footprint | :slim or :main-slim |
Features omitted from the slim image require external services. |
| NVIDIA acceleration for Open WebUI’s own local models | :cuda or :cuda126 with --gpus all |
Requires an NVIDIA driver, CUDA Container Toolkit and working Docker GPU support. |
| One container containing the interface and Ollama | :ollama with an Ollama volume |
Simpler topology, but model files still need persistent storage. |
| CPU-only Ollama | Standard :main plus host or container Ollama |
No GPU toolkit is needed; larger models may run more slowly, and the documentation provides no universal speed figure. |
Open WebUI documentation reports “197 MB instead of 1.62 GB to download, 88% smaller on Linux/amd64” for the slim image comparison. Those figures are project-published download figures for Linux/amd64, not a promise about every architecture or the memory required while models run.
Recommended Free Tools
How do you enable an NVIDIA GPU in Docker?
GPU acceleration is optional. It applies only when the relevant workload and image support CUDA; adding a GPU flag to a machine without a correctly configured NVIDIA runtime causes the container to fail rather than making CPU execution faster.
- Install a compatible NVIDIA driver on the host.
- Install the NVIDIA Container Toolkit and configure Docker according to its platform instructions.
- Verify that Docker can see the device with a small CUDA test container or your platform’s documented diagnostic command.
- Use an Open WebUI CUDA image, such as
ghcr.io/open-webui/open-webui:cudaor:cuda126, and add--gpus alltodocker run. In Compose, add the GPU device reservation supported by your Docker Compose version. - For an Ollama container, likewise add
--gpus=alland keep its model volume. If Ollama runs on the host, configure the host installation instead of adding a GPU flag to the Open WebUI container.
VRAM requirements vary by model size, quantization, context length and whether layers spill to system memory. No universal RAM or VRAM threshold or independent performance benchmark is established here, so select a model only after checking its own hardware guidance.
Also popular on Amazon
- 5-in-1 Connectivity: Equipped with a 4K HDMI port, a 5 Gbps USB-C data port, two 5 Gbps USB-A ports, and a USB C 100W PD-IN port. Note: The USB C 100W PD-IN port supports only charging and does not support data transfer devices such as headphones or speakers.
- Powerful Pass-Through Charging: Supports up to 85W pass-through charging so you can power up your laptop while you use the hub. Note: Pass-through charging requires a charger (not included). Note: To achieve full power for iPad, we recommend using a 45W wall charger.
- Transfer Files in Seconds: Move files to and from your laptop at speeds of up to 5 Gbps via the USB-C and USB-A data ports. Note: The USB C 5Gbps Data port does not support video output.
- HD Display: Connect to the HDMI port to stream or mirror content to an external monitor in resolutions of up to 4K@30Hz. Note: The USB-C ports do not support video output.
- What You Get: Anker 332 USB-C Hub (5-in-1), welcome guide, our worry-free 18-month warranty, and friendly customer service.
How do you keep chats, accounts and models safe during updates?
Separate the interface data from model storage and back up both before upgrades. The Open WebUI data is in the open-webui volume; an Ollama container normally stores models in ollama:/root/.ollama.
- Stop writes briefly if you need a consistent backup, then create an archive of the Open WebUI volume using a temporary container and Docker’s volume mount facilities.
- Back up the Ollama volume separately if redownloading models would be impractical.
- Pin an image tag or digest when reproducibility matters instead of relying indefinitely on a moving
:maintag. - Run
docker compose pullanddocker compose up -d(or recreate thedocker runcontainer with the same volume and secret). - Test login, chat history, endpoint connectivity and one model request before allowing normal use.
Changing the secret during a recreation can invalidate sessions. Removing volumes is a destructive operation; use docker compose down -v only when you deliberately intend to erase the Open WebUI data.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →What should you check when the interface cannot see a model?
- “Connection refused” or an empty model list: verify that Ollama is running, that port 11434 (or the documented runner port 12434) is listening, and that the URL is reachable from inside the Open WebUI container.
- Host Ollama works from the host but not from Docker: confirm the
host.docker.internal:host-gatewaymapping is present on Linux and thatOLLAMA_BASE_URLis not pointing at an obsolete address. - Separate containers cannot communicate: place them on the same user-defined Docker network and use the Ollama service name, or publish the port and use a reachable host address.
- Login disappears after every recreation: set a stable
WEBUI_SECRET_KEYand keep using the same Open WebUI data volume. - Chats vanish after an update: inspect the container’s mounts with
docker inspect open-webui. Recreating without-v open-webui:/app/backend/datastarts a new, empty data store. - GPU startup failure: remove GPU flags for a CPU test, then repair the host driver, NVIDIA Container Toolkit and Docker runtime before switching back to a CUDA image.
- Container repeatedly restarts: inspect
docker logs open-webui(ordocker compose logs open-webui) for an invalid environment variable, port collision, missing GPU runtime or endpoint configuration error.
Can a local Docker setup really replace a paid AI subscription?
It can remove recurring API charges for workloads that your own hardware can run, but “free” is a goal rather than a measured cost guarantee. You still provide electricity, storage, hardware and your own time for updates and troubleshooting; a remote service may remain preferable for very large models or high-volume workloads.
Start with a model that fits available memory, then measure response speed and quality for your actual prompts. The supplied documentation does not establish a universal performance result, subscription-price comparison, or minimum hardware specification.
What is the simplest architecture for your situation?
- One desktop and occasional use: run Ollama on the host and the standard Open WebUI image with the host-gateway mapping.
- A dedicated inference machine: run Ollama where the GPU and model storage live, set
OLLAMA_BASE_URLin Open WebUI, and restrict network access to trusted clients. - Docker-managed services: run Open WebUI and Ollama as separate Compose services on one network, with independent persistent volumes.
- Minimal download: choose a slim image only after confirming that the omitted speech-to-text or embedding features are available from services you can operate.
- All-in-one convenience: use the
:ollamaimage when a single container suits your workflow, while still backing up its model volume.
Whichever layout you choose, keep the Open WebUI volume, Ollama model storage and secret under your control. That combination—not the browser page alone—is what makes the installation reproducible after a restart or upgrade.
Frequently Asked Questions
Can I run Open WebUI on a Raspberry Pi or ARM computer?
The project lists Linux ARM64 support, including Raspberry Pi, as well as x86_64 Linux, macOS and Windows. Actual model capacity depends on the Pi or ARM machine’s memory and on whether your chosen endpoint provides a compatible build.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Do I need to expose Open WebUI to the public internet?
No. Binding port 3000 on your local machine is sufficient for devices that can reach that machine. If you publish it externally, add a reverse proxy with TLS, authentication and network restrictions rather than forwarding an unauthenticated Docker port.
What happens if I forget the Open WebUI secret?
You can recreate the container with a newly generated secret, but existing sessions may be invalidated. Keep a secure copy of the original value with your deployment configuration so routine upgrades do not cause avoidable logouts.
Can multiple people use one local Open WebUI installation?
Yes, the first account can be used to administer the instance and additional accounts can be created through the interface. Treat the endpoint and its data volume as shared infrastructure: access to the Docker host can expose stored chats and configuration.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
