DeepSeek-R1 on Ollama using Open-WebUI Docker Deployment

Deploying Open-WebUI with Ollama using Docker Compose (CPU-only setup) provides a streamlined way to run AI models locally without GPU hardware. This implementation uses a custom entrypoint script to download the deepseek-r1:8b model automatically when a container is launched. Here's an implementation guide:

📋 Prerequisites

• Docker Engine v20.10.10+
• Docker Compose v2.20.0+
• 8GB RAM minimum (16GB recommended for larger models)
• 20GB+ free disk space
• Linux/macOS/WSL2 (Windows)

🚀 Quick Start

1. Clone the repo

git clone https://github.com/ntalekt/deepseek-r1-docker-compose.git

2. Start Services

docker compose up -d

3. Verify Installation

docker compose ps

Expected Output:

NAME	COMMAND	SERVICE	STATUS	PORTS
ollama	"/bin/ollama serve"	ollama	running	0.0.0.0:11434->11434/tcp
open-webui	"bash start.sh"	open-webui	running	0.0.0.0:3000->8080/tcp

4. Check Logs

docker compose logs -f ollama
docker compose logs -f open-webui

5. Uninstall

docker compose down --volumes --rmi all

🌐 Access the Web UI

1. Open browser to http://localhost:3000
2. Create admin account
3. Start chatting!

Thinking

Answer

---

📄 License: MIT
❗ Note: CPU inference will be significantly slower than GPU-accelerated setups. For production use, consider GPU-enabled hardware.