OPENWEBUI

What is Open WebUI?

Open WebUI is an extensible, feature-rich, self-hosted AI platform designed to operate entirely offline. It runs on various environments (Docker, pip, Kubernetes, desktop app) and supports multiple LLM runners including Ollama and OpenAI-compatible APIs, with built-in support for connecting external services through MCP, MCPO, and OpenAPI tool servers.

How to use Open WebUI?

Install Open WebUI via Python pip (pip install open-webui) or Docker (e.g., docker run -d -p 3000:8080 ...), then start the server with open-webui serve. After setup, connect to Ollama or any OpenAI-compatible API, and enable MCP support by configuring filters, actions, pipes, or tools within the plugin system.

Key features of Open WebUI

• Effortless setup via pip, uv, Docker, or Kubernetes
• Broad model integration with Ollama and any OpenAI‑compatible API
• Plugin support for MCP, MCPO, and OpenAPI tool servers
• Local RAG with hybrid search and multiple vector databases
• Real‑time team collaboration through Channels and persistent memory
• Multi‑model conversations and built‑in image generation/editing

Use cases of Open WebUI

• Deploy a fully offline AI assistant for document analysis and Q&A using local LLMs and RAG
• Build custom agents with tailored instructions, tools, and knowledge bases for team workflows
• Enable real‑time collaboration by creating shared Channels where AI and team members interact simultaneously
• Automate recurring prompts (e.g., daily reports) with calendar integration and on‑schedule runs
• Extend functionality through MCP servers, connecting external APIs, databases, and services

FAQ from Open WebUI

What is Open WebUI?

Open WebUI is a self-hosted AI platform that runs offline, supports multiple LLM runners (Ollama, OpenAI‑compatible APIs), and includes built‑in RAG, image generation, and extensibility via plugins and MCP.

How do I install Open WebUI?

You can install it using Python pip (pip install open-webui) or Docker (e.g., docker run -d -p 3000:8080 -v open-webui:/app/backend/data ghcr.io/open-webui/open-webui:main). After installation, run open-webui serve and access it at http://localhost:8080.

Does Open WebUI support MCP?

Yes. Open WebUI supports MCP, MCPO, and OpenAPI tool servers through its plugin system (Filters, Actions, Pipes, Tools, and Skills) for connecting external services.

Can Open WebUI run entirely offline?

Yes. Open WebUI is designed to operate offline, with a built‑in inference engine for RAG and support for local LLM runners such as Ollama. It requires no internet connection once installed and configured.

What database and storage options are available?

You can choose SQLite (with optional encryption) or PostgreSQL for metadata, and store files locally or on S3, Google Cloud Storage, or Azure Blob Storage. Vector databases include ChromaDB, PGVector, Qdrant, Milvus, and others.