Open WebUI

Open WebUI is a self-hosted AI platform with 156,000+ stars on GitHub and hundreds of millions of downloads that gives you a polished, ChatGPT-style interface for models you control.

Overview

Open WebUI is an extensible, feature-rich web interface that works entirely offline on your own infrastructure. It connects to Ollama and any OpenAI-compatible API — OpenAI, Anthropic and many others — so local and cloud models live side by side in one place. Around the chat core it layers retrieval-augmented generation over your documents, voice and video calls, image generation, multi-user access with roles and permissions, and a Python-based system of tools, functions and pipelines for extending behavior. A community of more than half a million members shares prompts, models and extensions. The project ships frequent releases — v0.12 arrived in October 2026 with realtime voice calls and channel apps — and it scales from a single laptop (installable with pip) to enterprise deployments with SSO, RBAC and audit logs.

Key Features

• Any model, one interface – Ollama plus OpenAI-compatible APIs including OpenAI and Anthropic.
• Built-in RAG – chat with documents using local or cloud embedding and retrieval pipelines.
• Voice, video and images – calls, speech and image generation integrated into the chat experience.
• Multi-user platform – accounts, groups, RBAC, SSO and audit logging for teams and enterprises.
• Extensible by design – tools, functions, pipelines and filters written in Python.
• Truly self-hosted – runs offline via Docker or pip; your data stays on your machines.

💳 Pricing

Open WebUI is free to self-host and open source. Enterprise licensing with support is available for organizations that need it — see openwebui.com for details.

Who Is It For

Individuals who want a private ChatGPT alternative, and teams or enterprises that need a governed, self-hosted AI front end over local and cloud models.