Hi everyone,
Open WebUI is a self-hosted, open-source AI platform designed for ease of use and extensibility. It offers a range of features for interacting with a variety of LLMs:
🔗 Broad LLM Compatibility: Works with Ollama and OpenAI-compatible APIs (e.g., LMStudio, Mistral), plus built-in RAG, web browsing/search, and image generation.
✨ Highly Extensible: Includes a model builder, Python function calling, and a plugin system.
🔒 Secure & User-Friendly: Role-Based Access Control, responsive design (with PWA), and full Markdown/LaTeX support.
🚀 Easy Deployment: Docker and Kubernetes support for simplified installation.
Open WebUI is designed for offline use. I know this is crucial for many people.
Open WebUI is the self-hosted chat front-end I actually kept running. Point it at Ollama and you get a clean multi-model UI, RAG over your own docs, and per-user access without sending anything to a cloud. Setup beyond the basic compose takes some patience, and it gets heavy with several models loaded at once. For a private, offline-capable ChatGPT-style UI it is the best I have tried.
Flowtica Scribe