Skip to content
buildbay.
Open WebUI logo

Open WebUI

open-webui · AI assistants

GitHub stars

153,759
—Collecting 30d data

Stats updated

153,345 stars on Sep 27 to 153,759 stars on Oct 2, up 414.6 measured star snapshots

What is Open WebUI?

Open WebUI turns a model endpoint into a complete AI workspace. It can run entirely on private infrastructure, connect to local or hosted models, and layer retrieval, search, tools, permissions, and collaboration onto a consistent chat interface that works across desktop and mobile devices.

Open WebUI is a self-hosted AI platform for working with Ollama, OpenAI-compatible APIs, and other model providers through one responsive interface. It adds knowledge retrieval, web search, tools, permissions, voice features, and extensible pipelines around the core chat experience.

Core capabilities

  • Connects Ollama, OpenAI-compatible APIs, and multiple model providers in one self-hosted interface.
  • Adds retrieval over documents, web search, tools, image generation, voice, and model-specific workflows.
  • Supports user accounts, groups, granular permissions, responsive access, and offline-first deployments.

One interface for many model backends

Open WebUI works with Ollama and OpenAI-compatible APIs, so a deployment can expose local models, commercial providers, or internal gateways through the same interface. Users can compare models and keep their familiar workspace while administrators control which connections are available.

  • Ollama
  • OpenAI-compatible APIs
  • local models
  • self-hosted

Knowledge, tools, and team controls

Built-in retrieval can ground responses in documents and knowledge collections, while web search, functions, pipelines, and MCP integrations extend what assistants can do. User groups and granular permissions make those capabilities manageable in a shared deployment.

  • RAG
  • web search
  • tools
  • permissions

Where it fits

Use cases

  1. 01

    Private, self-hosted AI chat

    Run a familiar AI chat interface on infrastructure you control, connect local Ollama models or remote APIs, and keep conversations and configuration inside your deployment.

  2. 02

    Multi-model team workspace

    Give teams one responsive workspace for different model providers while administrators manage users, groups, permissions, models, and shared resources.

  3. 03

    Knowledge and tool-enabled assistants

    Ground answers in uploaded knowledge, search the web, call tools and functions, generate images, and extend workflows through pipelines and community integrations.