Open WebUI

Open WebUI is a self-hosted chat workspace for individuals and teams using local LLMs and cloud models. It can work entirely offline with local models.

• Model connections: Chat through Ollama or OpenAI-compatible backends such as LM Studio and vLLM, or use providers including OpenAI, Anthropic and OpenRouter.
• Chat with context: Use retrieval, search, voice and vision, and carry facts about yourself across conversations with persistent memory.
• Agents and tools: Combine models with instructions, knowledge and tools, add Python extensions, and connect MCP or OpenAPI tool servers.
• File tasks: Give agents access to Open Terminal to analyze data, run scripts and return generated files directly in chat.
• Shared workspace: Draft material in Notes and attach it to conversations, collaborate with people and models in Channels, and manage access through roles and user groups.

The interface runs on your own computer or server, including through Docker, and doesn't require an external account to get started. Cloud model use sends requests to the selected provider. Community libraries supply prompts, model presets, tools and functions.