Libre WebUI / Documentation

Make whatever comes next.

Run Libre WebUI locally, bring the models you choose, work with documents and media, or give an agent a durable project workspace. Everything opens as a tab in one workspace. Start with the task you want to accomplish.

Install Libre WebUI Try the demo View source

One-command local start

npx libre-webui

Open your private workspace at http://localhost:8080.

01 / Choose a task

What do you want to do?

Run privately

Install Libre WebUI on your own machine and begin with Ollama or another model you control.

Follow the Quick Start

Connect your models

Use bundled providers, an OpenAI-compatible service, a self-hosted gateway, or a coding agent already installed on the machine.

Connect a model provider

Build with Work

Give an agent a durable project with conversation, files, a terminal, tools, and a live application preview.

Explore Libre WebUI Work

Deploy for others

Move from a local workspace to a maintained deployment with persistent data and access controls.

Plan a deployment

02 / Get around

Everything is one tab away.

Start from Home

Home greets you, offers what to start, and lists the chats and Work sessions worth picking back up, with a live badge when a Work runtime is active.

Keep work side by side

Chats, Work sessions, and pages open as tabs beside Home. They survive a reload, so returning to a project returns you to where you left it.

Search instead of hunting

Press Cmd/Ctrl + K anywhere to reach a chat, a Work session, or an action. It works while you are typing a message, so a draft is never lost to navigation.

All keyboard shortcuts

03 / Browse guides

Find the path that matches your work.

Models and providers

Choose where inference runs and connect the services that fit your work.

Create and automate

Turn conversations into durable projects, reusable context, and working material.

Deploy and administer

Run a maintained installation with deliberate access, data, and configuration choices.