Edge Rewrite
// HTMLRewriter · presentation

This page was redesigned at the edge.

Cloudflare fetched the original article and streamed it through HTMLRewriter to apply an entirely new visual system without rebuilding the source page.

// request.cf · coarse context

A page that knows where it met you.

Only coarse request metadata is shown. This demo does not display or persist visitor IP addresses.

Country
US
Cloudflare location
CMH
Connection
HTTP/2
Language
Not provided

Ray ID: a3fe98abec668043

Jump to content

// Workers AI · dad joke modeWhat did Open WebUI say to the user? "You can open up to me".

From Wikipedia, the free encyclopedia

Open WebUI
Original authorTimothy J. Baek
DeveloperOpen WebUI
Release2023
Written inPython, JavaScript, TypeScript
TypeLarge language model interface
LicenseOpen WebUI License
Websiteopenwebui.com
Repositoryhttps://github.com/open-webui/open-webui

Open WebUI is a self-hosted graphical user interface for interacting with artificial intelligence models, particularly large language models. It can be used as a browser-based interface for locally hosted models through software such as Ollama and for remote models provided through OpenAI-compatible application programming interfaces (APIs).[1][2]

When used with locally hosted models, Open WebUI can operate entirely offline, while remotely hosted models can be accessed over the Internet through third-party providers. It provides a ChatGPT-style interface while allowing users to select among different models and providers.[3]

History

[edit]

Open WebUI was created by Timothy J. Baek in 2023 and was originally named Ollama WebUI. The project initially served as a web interface for Ollama before expanding to support additional model providers and APIs. It was renamed Open WebUI in February 2024 to reflect its broader scope.[4]

Features

[edit]

Open WebUI provides a web-based chat interface that can connect to Ollama and services implementing an OpenAI-compatible API. Users can interact with multiple models from the same interface and deploy the software using methods including Docker, Kubernetes, or Python.[5][6]

The software includes support for retrieval-augmented generation (RAG) using uploaded documents and knowledge bases. Other features include web search, image generation, voice interaction, model tools and external extensions.[7]

Open WebUI can connect to external tools and services through mechanisms including OpenAPI servers and the Model Context Protocol (MCP). It also supports Python-based tools that execute within an Open WebUI installation.[8]

Open WebUI is primarily used with locally hosted large language models, but it can also access models hosted by third-party providers through compatible APIs.[9]

Licensing

[edit]

Early versions of Open WebUI were released under the MIT License, and the project later changed to the BSD 3-Clause License. Beginning with version 0.6.6 in April 2025, new code has been distributed under the Open WebUI License, which generally permits modification and redistribution but restricts removal or alteration of Open WebUI branding in deployments with more than 50 users. The project states that the license does not meet the Open Source Initiative's definition of an open-source license.[14]

Code released under the project's earlier licenses remains governed by those licenses. The project's license notice states that earlier portions of the codebase remain under the MIT and BSD 3-Clause licenses, while later contributions are governed by the Open WebUI License.[15]

Security

[edit]

In 2025, a vulnerability in Open WebUI's Direct Connections feature, designated CVE-2025-64496, was disclosed. The vulnerability could allow malicious model servers to execute JavaScript in a user's browser and potentially lead to account takeover or remote code execution when combined with other functionality. It affected version 0.6.34 and earlier and was patched in version 0.6.35.[16]

See also

[edit]
[edit]

References

[edit]
  1. Devine, Richard (August 5, 2025). "Ollama Launches User-Friendly GUI App for Effortless Local AI on Windows 11 — Here's What You Need to Know". Windows Central. Retrieved September 12, 2026.
  2. "Open WebUI". vLLM Documentation. vLLM. Retrieved September 12, 2026.
  3. Fadilpašić, Sead (January 6, 2026). "This WebUI vulnerability allows remote code execution – here's how to stay safe". TechRadar. Retrieved September 12, 2026.
  4. Baek, Timothy J. (January 30, 2024). "Announcing a New Chapter for Ollama-webui: Help Us Choose Our New Name!". GitHub. Open WebUI. Retrieved September 12, 2026.
  5. "Open WebUI". Open WebUI Documentation. Open WebUI. Retrieved September 12, 2026.
  6. "Developing Open WebUI". Open WebUI Documentation. Open WebUI. Retrieved September 12, 2026.
  7. "What You Can Do with Open WebUI". Open WebUI Documentation. Open WebUI. Retrieved September 12, 2026.
  8. "Tools". Open WebUI Documentation. Open WebUI. Retrieved September 12, 2026.
  9. "Connect a Provider". Open WebUI Documentation. Open WebUI. Retrieved September 12, 2026.
  10. "Open WebUI integration". Docker Docs. Docker. Retrieved September 12, 2026.
  11. "Open WebUI". Lemonade Server Documentation. Retrieved September 12, 2026.
  12. "Try LocalAI". LocalAI Documentation. LocalAI. Retrieved September 12, 2026.
  13. "HTTP Model Server". GitHub. MLX. Retrieved September 12, 2026.
  14. "Open WebUI License". Open WebUI Documentation. Open WebUI. Retrieved September 12, 2026.
  15. "LICENSE_NOTICE". GitHub. Open WebUI. Retrieved September 12, 2026.
  16. Fadilpašić, Sead (January 6, 2026). "This WebUI vulnerability allows remote code execution – here's how to stay safe". TechRadar. Retrieved September 12, 2026.