Open-WebUI

Open WebUI provides a self‑hosted web interface for interacting with large language models. It can connect to locally run Ollama instances as well as any OpenAI‑compatible API, allowing users to switch between on‑premise and cloud‑based models without changing the front end. The platform includes built‑in tools for Retrieval‑Augmented Generation, markdown and LaTeX rendering, and a progressive web app that works on desktop and mobile browsers.

The system is designed for extensibility: administrators can define granular user roles and permissions, and developers can add custom Python functions or create new Ollama models directly from the UI. Integrated voice and video call support, along with multiple speech‑to‑text and text‑to‑speech providers, enable hands‑free interaction. Persistent key‑value storage and support for nine vector databases facilitate document loading, search, and collaborative workflows.

Open WebUI is released under the BSD‑3‑Clause license, can be deployed via Docker or Kubernetes, and targets users who want full control over their AI stack—ranging from individual hobbyists to organizations that require on‑premise data protection and customizable tooling.