Getting Started with Open WebUI
From zero to your first AI conversation in under five minutes.
Open WebUI runs anywhere (Docker, Kubernetes, pip, bare metal) and connects to Ollama, OpenAI-compatible, and Open Responses providers out of the box. Pick an install method, connect a provider, and start chatting.
โฑ๏ธ Quick Startโ
Install Open WebUI, connect a model, and start chatting.
Everything you need for a working setup. Choose Docker for the fastest path, Python for lightweight installs, or Kubernetes for production orchestration. Each guide includes connecting your first model provider.
| ๐ณ Docker | One-command deploy, the officially recommended path |
| ๐ฆ Python (pip / uv) | Lightweight install for low-resource or manual setups |
| โธ๏ธ Kubernetes (Helm) | Production-ready orchestration with scaling |
| ๐ฅ๏ธ Desktop app | Native app, no Docker required |
| ๐ Connect a provider | Ollama, OpenAI, Anthropic, llama.cpp, vLLM, and more |
| โ๏ธ Understanding settings | Learn how Admin Settings and User Settings work together |
Essentials for Open WebUIโ
Installed and chatting โ now what?
Five short sections that cover the things every new user eventually wishes they'd known on day one: what plugins are and how to install them, why long chats eventually error out (and how to fix it with a filter), the "invisible" Task Model that powers titles/autocomplete, getting started with RAG over your own documents, and turning on Native tool calling.
| ๐งฉ Plugins | Tools, Pipes, Filters, Actions โ the extensibility story |
| ๐ง Context management | Why long chats hit a wall and how to handle it |
| ๐ค Task models | Keep titles, tags, and autocomplete off your main model |
| ๐ Basic RAG | Chatting with your own documents |
| ๐ง Tool calling | Native mode + first Tools to install |
๐ค Connect an Agentโ
Go beyond simple model providers. Connect an autonomous AI agent.
AI agents like Hermes Agent and OpenClaw bring their own tools (terminal, file ops, web search, memory) and use Open WebUI as a rich chat frontend. The agent decides when to use tools, executes them, and streams results back to you.
| ๐ง Hermes Agent | Nous Research's agent with terminal, file ops, web search, and skills |
| ๐พ OpenClaw | Self-hosted agent framework with shell access, web browsing, and channel bots |
Sharing Open WebUIโ
Bring AI to your entire organization with a single deployment.
Open WebUI isn't just a local interface for AI. It is designed to be a centralized AI operating system for teams. Deploy it once to leverage streamlined onboarding, collaborative intelligence, resource pooling, and centralized security.
| LAN & Tunnels | Tailscale, Cloudflare Tunnels, and local IP access |
| Reverse Proxies | Secure your instance with Nginx, Caddy, or HAProxy |
| Team Onboarding | Admin approval flows and Enterprise SSO integrations |
| Shared Context | Channels, shared chats, and common knowledge bases |
Learn how to share Open WebUI โ
๐ ๏ธ Advanced Topicsโ
Scale, observe, and customize your deployment.
Go beyond the basics. Configure environment variables, connect external databases, add cloud storage, enable OpenTelemetry, or scale horizontally with Redis and multiple workers.
| โ๏ธ Scaling | Multi-worker, multi-node, Redis-backed sessions |
| ๐ Logging & observability | OpenTelemetry traces, metrics, and structured logs |
| ๐งช Development setup | Run from source for local development and testing |
๐ Updatingโ
Stay current with the latest features and security patches.
Update manually with a single Docker or pip command, or automate with Watchtower, WUD, or Diun. Includes backup/restore procedures and version pinning for production.
| ๐ Manual update | Docker pull + recreate, or pip upgrade |
| ๐ค Automated updates | Watchtower, WUD, or Diun |
| ๐ Version pinning | Lock to a specific release for stability |
| ๐พ Backup & restore | One-command volume backup and recovery |