# Open WebUI

Self-hosted web interface for interacting with LLMs (Ollama, OpenAI-compatible APIs, etc.).

Includes:
- **Open WebUI** — Main chat interface
- **Open Terminal** — Browser-based terminal
- **Apache Tika** — Document parsing for RAG

## Setup

1. Copy `.env.example` to `.env` and set your API key:
   ```
   cp .env.example .env
   ```

2. Start:
   ```
   docker compose up -d
   ```

## Access

| Service       | URL                     |
|---------------|--------------------------|
| Open WebUI    | `http://<host>:3000`    |
| Open Terminal | `http://<host>:8001`    |
| Tika          | `http://<host>:9998`    |

## What to Configure

- **`.env`** — Set `OPEN_TERMINAL_API_KEY` to a secure random string.
- **Ollama / LLM backend** — The `host.docker.internal` extra host is configured to reach a local Ollama instance. If your LLM backend runs elsewhere, update the host or add environment variables like `OPENAI_API_BASE_URL`.
- **RAG settings** — `RAG_EMBEDDING_ENGINE=openai` uses OpenAI-compatible embeddings. Change if using a different engine.
- **Persistent volumes** — `open-webui` and `open-terminal` are named Docker volumes. Data persists across container restarts.