open-webui
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
What is open-webui?
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
Overview of open-webui
What it is
Open WebUI is a self-hosted AI platform described as a home for AI. It is extensible, feature-rich, user-friendly, and built to run entirely offline. It supports Ollama and OpenAI-compatible APIs, providing a provider-agnostic interface for local and cloud-based models. Metadata lists it under AI and marks it as open source.
Key features
- Setup via pip, uv, Docker, or Kubernetes, with
:ollamaand:cudatagged images for container deployments. - Broad model and API integration: connect OpenAI-compatible APIs and local Ollama models, including LMStudio, GroqCloud, Mistral, OpenRouter, vLLM, and more.
- Granular RBAC and user groups, with roles, groups, and permissions defined by administrators.
- Plugin support through Filters, Actions, Pipes, Tools, and Skills, plus MCP, MCPO, and OpenAPI tool servers.
- Models and agents with custom instructions, tools, knowledge, dynamic variables, and per-user/group access control.
- Local RAG with 9 vector databases, multiple content-extraction engines, hybrid search, reranking, and full-context mode.
- Web search, web browsing, image generation and editing, multi-model conversations, voice/video calls, Notes, Channels, persistent memory, Calendar, Automations, usage analytics, and enterprise authentication.
Use cases
Use Open WebUI to run a self-hosted AI interface for local and cloud models, mix providers through OpenAI-compatible APIs, build specialized agents, give agents terminal and filesystem access for multi-step tasks, collaborate in Channels, manage schedules with AI, and deploy with RBAC, SSO/LDAP, SCIM, observability, and horizontal scalability. It supports team and enterprise scenarios through user groups, shared calendars, Channels, and Terminals (Enterprise).
Getting started
The README states installation is possible via pip, uv, Docker, or Kubernetes (kubectl, kustomize, or helm), with :ollama and :cuda tagged images for container deployments. It does not provide exact commands; see the repository README.
License and pricing
Metadata labels the project open source and lists the license as NOASSERTION; the README does not state a specific license. The README mentions an Enterprise Plan and a sales team, but does not provide pricing.
Based on the project's public README and repository metadata.