openwebui

name: openwebui description: Open WebUI self-hosted AI chat interface. This skill should be used when deploying (Docker/Helm/pip), configuring auth/SSO/RBAC, RAG/vector DBs, web search, plugins/MCP, scaling, or reverse proxy.

Open WebUI

Self-hosted AI chat interface supporting Ollama, OpenAI-compatible APIs, and custom pipelines. Provides RAG, web search, code execution, image generation, TTS/STT, RBAC, SSO, and extensibility via plugins/MCP.

Quick Start

# Docker with Ollama
docker run -d -p 3000:8080 -v open-webui:/app/backend/data \
  -e OLLAMA_BASE_URL=http://host.docker.internal:11434 \
  --name open-webui ghcr.io/open-webui/open-webui:main

# pip
pip install open-webui && open-webui serve

Reference Index

Load relevant reference files based on the task. Files are organized by topic area.

Environment Variables (split by category)

• env-configuration.md -- Index pointing to split files
• env-general.md -- App config, directories, logging, AIOHTTP, streaming
• env-auth.md -- OAuth, OIDC, SSO, security, cookies, API keys
• env-rag.md -- RAG, embedding, content extraction, web search, document processing
• env-connections.md -- Ollama, OpenAI, database, Redis, WebSocket, vector DBs
• env-media.md -- Audio STT/TTS, Whisper, image generation engines
• env-features.md -- Tasks, code execution, autocomplete, permissions

Authentication & Security

• auth-sso.md -- OAuth (Google/Microsoft/GitHub/OIDC), session management, role/group mapping
• auth-sso-trusted-header.md -- Trusted header auth, Tailscale, Cloudflare, oauth2-proxy, Authentik, Authelia
• auth-ldap.md -- OpenLDAP Docker setup, LDIF seeding, env vars, troubleshooting
• auth-scim.md -- SCIM 2.0 provisioning, user/group operations, filtering, Okta setup
• rbac.md -- Roles (admin/user/pending), groups, ACLs, headless admin
• rbac-permissions.md -- All permission categories (workspace, sharing, chat, features, settings)
• rbac-api-keys.md -- API key setup, generation, usage examples

Administration

• administration.md -- Banners config, webhook types (admin/user/channel)
• administration-analytics.md -- Dashboard, token tracking, cost estimation, API endpoints
• administration-evaluation.md -- Arena model, ELO leaderboard, chat snapshots

Chat Features

• chat-features-overview.md -- Autocomplete, follow-up prompts, message queue
• chat-features-params.md -- Chat parameters hierarchy, URL parameters
• chat-features-sharing-history.md -- Chat sharing, history search
• chat-features-multimodel-queue.md -- Multi-model chats, conversation organization/folders

Code Execution

• code-execution-overview.md -- Backends comparison (Pyodide/Jupyter/Terminal/Terminals)
• code-execution-python.md -- Python code interpreter, Pyodide libraries, inline images
• code-execution-artifacts-mermaid.md -- Artifacts, MermaidJS rendering

Reasoning & Temporal

• reasoning-models.md -- Thinking tags, reasoning_tags config, provider compatibility
• reasoning-models-troubleshooting.md -- Ollama reasoning parser, Anthropic workarounds, FAQ

Data Controls

• data-controls.md -- Shared/archived chats, file management
• data-controls-import-export.md -- JSON schema, ChatGPT migration, message tree
• memory-connections.md -- Memory tools, direct connections, CORS

RAG & Document Processing

• rag.md -- Chunking, markdown splitting, chunk min size
• rag-part2.md -- Embedding, re-indexing, KV cache, Google Drive, YouTube
• document-extraction.md -- Extraction overview, supported formats
• agentic-search.md -- Native vs traditional RAG, search_web/fetch_url tools, deep research

Media Generation

• image-generation.md -- Native tool-based generation, inpainting, compositing
• speech-to-text.md -- STT config, all env vars (Whisper/OpenAI/Azure/Deepgram)
• speech-to-text-mistral.md -- Mistral Voxtral integration
• text-to-speech.md -- OpenAI TTS, voices, per-model voice, Azure

Plugins & Extensibility

• plugin-overview.md -- Architecture (Tools vs Functions vs Pipelines)
• plugin-functions.md -- Function types overview (Pipe/Filter/Action)
• plugin-functions-action.md -- Action class, multi-actions, priority
• plugin-functions-filter.md -- Filter skeleton, toggle filters, two-tier system
• plugin-functions-filter-examples.md -- Stream hook, rate limiting, best practices
• plugin-functions-pipe.md -- Pipe/manifold pattern, OpenAI proxy, internal functions
• plugin-tools.md -- Tool taxonomy, calling modes, built-in system tools
• plugin-tools-development.md -- Tool class structure, OAuth tokens, citations
• plugin-development-events.md -- All event types, persistence, external tool events
• plugin-development-valves.md -- Valves/UserValves, password/select/dynamic inputs
• plugin-development-rich-ui.md -- HTMLResponse, iframe, sandbox, prompt submission
• openapi-servers.md -- OpenAPI tool servers, mcpo proxy, CORS
• mcp.md -- MCP native support, auth modes, troubleshooting
• pipelines.md -- Pipelines Docker setup, filter/pipe types, valves
• plugin-migration.md -- 0.4 to 0.5 migration guide

Workspace

• workspace-models.md -- Model creation, capabilities, global defaults
• workspace-prompts.md -- Slash commands, custom input types, versioning
• workspace-knowledge.md -- Retrieval modes, agentic tools, API endpoints
• workspace-skills.md -- Markdown skills, lazy-loading, import/export
• notes.md -- Rich editor, AI enhance, agentic note tools
• channels.md -- Real-time channels, @model tagging, setup

Open Terminal

• open-terminal.md -- Overview, installation, connecting, file browser
• open-terminal-advanced.md -- Configuration, multi-user, security
• terminals.md -- Enterprise terminals, Docker/Kubernetes backends
• terminals-policies.md -- Terminal policies configuration

Installation & Deployment

• install-docker.md -- Docker run, image variants, GPU, uninstall
• install-docker-compose-swarm.md -- Docker Compose, Swarm stack
• install-docker-podman-wsl.md -- Podman, Quadlets, WSL
• install-python.md -- pip, uv, venv, Conda
• install-kubernetes.md -- Helm chart setup
• llm-providers.md -- Ollama, OpenAI, Azure, Anthropic
• llm-providers-compatible.md -- Gemini, DeepSeek, Mistral, Groq, LiteLLM
• llm-providers-local.md -- vLLM, Llama.cpp, Open Responses
• scaling.md -- PostgreSQL, Redis, multiple instances, vector DBs
• scaling-storage.md -- Shared storage, content extraction, observability
• updating.md -- Update strategies, rollback, Watchtower

Reference & Infrastructure

• api-endpoints.md -- Chat completions, Anthropic API, RAG, filters
• https-reverse-proxy.md -- Cloudflare Tunnel, ngrok, Tailscale, Caddy, HAProxy
• nginx-configs.md -- NPM, Let's Encrypt, self-signed, Windows
• monitoring.md -- Health checks, Uptime Kuma, OpenTelemetry
• logging.md -- Log levels, JSON format, structured logging
• settings.md -- Admin vs User settings architecture
• development.md -- Local dev setup (SvelteKit + FastAPI)

Troubleshooting

• troubleshooting.md -- Connection, CORS, streaming, Ollama, password reset, RAG, SSO, audio, web search, image gen, scaling, performance, database migration

Enterprise

• enterprise.md -- Overview, licensing, integration, white-labeling
• enterprise-security.md -- Compliance (SOC2/HIPAA/GDPR), HA architecture
• enterprise-deployment.md -- Infrastructure requirements, env vars
• enterprise-deployment-patterns.md -- VM/pip, container service
• enterprise-deployment-kubernetes.md -- Helm values, HPA, updates

Tutorials & FAQ

• tutorials-auth.md -- Dual OAuth, Tailscale quick start
• tutorials-auth-tailscale.md -- Tailscale SSO, Docker Compose
• tutorials-integrations.md -- OneDrive/SharePoint business
• tutorials-integrations-continued.md -- OneDrive personal, DeepSeek R1
• faq.md -- Support, Docker networking, GPU, updating
• faq-continued.md -- STT/TTS, RAG context, API, MCP, scalability
• sharing.md -- Team sharing, LAN/Tailscale/Cloudflare, onboarding