Open WebUI
4 Intel
Open WebUI is an open-source, self-hosted interface for running large language models locally or via API, supporting backends such as Ollama and OpenAI-compatible providers. The project ships as a Docker- or Kubernetes-deployable web application with built-in retrieval-augmented generation support, and its GitHub organization maintains over two dozen related repositories, including a native desktop app. It is released under the Open WebUI License, which requires preserving the project's branding in derivative works.
Intel
Open WebUI v0.10.2 adds streamed reasoning display, so models emitting thinking content show it as it arrives and render it correctly in the chat overview and in exported conversations.
Open WebUI 0.13.0 adds a tabbed chat workspace for files, terminals and previews, and an Open Terminal Desktop tab that lets models and users share control of a desktop for computer-use tasks.
Open WebUI 0.12.4 adds optional lazy loading of skills, so full instructions are sent only when needed, and shows context compaction inline with expandable summaries. Compaction now also runs in temporary and short chats.
Open WebUI 0.12.5 adds an admin analytics Sessions tab showing per-chat tokens, errors, and tool-call timing, plus multi-page notes with live presence and a rename option in the chat menu.