Ollama and WebUI Installation on Arch Linux (Docker)

Ollama and WebUI Installation on Arch Linux (with Docker) 🤖#

Ollama is a modern platform that allows you to manage large language models (LLM) running locally. Open WebUI provides a visual interface to easily use these models through your browser. In this article, you will learn step-by-step Ollama and Open WebUI installation using Docker on Arch Linux, including all methods including GPU acceleration, single container bundle installation, and automatic updates.

💡 If Docker is not installed on your system, first check our Docker Installation on Arch Linux guide.

🔑 WebSocket Support#

WebSocket connection support is required for Open WebUI to function properly. Make sure your network configuration does not block WebSocket connections.

1️⃣ Installation Prerequisites#

• Arch Linux (up-to-date)
• Docker installed and running
• Internet connection
• Minimum 8 GB RAM (16 GB+ recommended for larger models)

• Ollama application:

• If Ollama is not installed on your system, install it with the following command:

bash yay -Syyu ollama

• If you want to run Ollama inside Docker as well, see the single container bundle installation section.

2️⃣ Scenario 1: Ollama Installed on Computer + WebUI via Docker#

In this scenario, Ollama is installed on your computer, and WebUI runs inside Docker.

CPU Installation#

docker run -d \
  -p 3000:8080 \
  --add-host=host.docker.internal:host-gateway \
  -v open-webui:/app/backend/data \
  --name open-webui \
  --restart always \
  ghcr.io/open-webui/open-webui:main

Installation with NVIDIA GPU Support#

docker run -d \
  -p 3000:8080 \
  --gpus all \
  --add-host=host.docker.internal:host-gateway \
  -v open-webui:/app/backend/data \
  --name open-webui \
  --restart always \
  ghcr.io/open-webui/open-webui:cuda

3️⃣ Scenario 2: Ollama + WebUI Single Container (Bundle Installation)#

With this method, both Ollama and Open WebUI are installed with a single command.

GPU Supported Installation#

docker run -d \
  -p 3000:8080 \
  --gpus=all \
  -v ollama:/root/.ollama \
  -v open-webui:/app/backend/data \
  --name open-webui \
  --restart always \
  ghcr.io/open-webui/open-webui:ollama

CPU Installation#

docker run -d \
  -p 3000:8080 \
  -v ollama:/root/.ollama \
  -v open-webui:/app/backend/data \
  --name open-webui \
  --restart always \
  ghcr.io/open-webui/open-webui:ollama

✅ This method runs both applications at once without dependency and configuration concerns.

4️⃣ NVIDIA GPU Support (Detailed)#

To achieve high performance with NVIDIA GPU:

sudo pacman -S nvidia nvidia-utils nvidia-container-toolkit
sudo nvidia-ctk runtime configure --runtime=docker
sudo systemctl restart docker

To test the installation:

nvidia-smi

5️⃣ AMD GPU (ROCm) Support#

AMD GPU users can enable acceleration with ROCm:

sudo pacman -S rocm-opencl-runtime rocm-hip-runtime

When running Docker:

--device=/dev/kfd --device=/dev/dri

add these parameters.

6️⃣ Accessing WebUI#

After installation, access from your browser:

http://localhost:3000

You will be asked to create a user account on first startup.

7️⃣ Updates#

Manual Update#

docker run --rm \
  -v /var/run/docker.sock:/var/run/docker.sock \
  containrrr/watchtower \
  --run-once open-webui

Automatic Update (Every 5 Minutes)#

docker run -d \
  --name watchtower \
  --restart unless-stopped \
  -v /var/run/docker.sock:/var/run/docker.sock \
  containrrr/watchtower \
  --interval 300 open-webui

🔧 Note: Write your own container name instead of open-webui.

8️⃣ Common Issues#

• WebUI cannot connect to Ollama: Use the single container bundle or don't forget the --add-host parameter in host connection.
• Slow model download: Depends on your internet speed. You can copy previously downloaded models into the ollama-data volume.
• Port conflict: Change ports like -p 3000:8080.

🔚 Conclusion#

In this guide, you learned:

• Ollama and Open WebUI installation on Arch Linux
• CPU, GPU, and single container bundle scenarios
• Update methods
• WebSocket requirements

💬 Which installation method did you prefer? Share your experiences!