LobeChat (81,314 Stars) 2026: Self-Host a Multi-Provider AI Assistant with Plugin Support

LobeChat (81,314 stars) is a modern, self-hostable AI chat interface with multi-provider support, plugins and a plugin marketplace. Here is the complete setup guide.

💡 What You Will Learn

LobeChat (81,314 stars) is a modern, self-hostable AI chat interface with multi-provider support, plugins and a plugin marketplace. Here is the complete setup guide.

## The short answer **lobehub/lobe-chat** (81,314 stars, TypeScript) is an open-source AI chat framework that you can deploy in one click. It supports OpenAI, Claude, Gemini, and dozens of providers (including OpenAI-compatible local endpoints), plus a rich plugin system that lets assistants call APIs, search the web, and render charts. ## What makes LobeChat popular - **Multi-provider**: switch models from the UI - no code changes - **Plugin marketplace**: install tools (web search, image generation, calculators) - **Agent store**: shareable assistant personas with custom system prompts - **File upload**: OCR, embeddings and RAG for your documents - **Beautiful UI**: modern design that feels like a polished consumer app ## Deploy with Docker ```bash docker run -d -p 3210:3210 \n -e OPENAI_API_KEY=sk-... \n --name lobe-chat \n lobehub/lobe-chat ``` Open http://localhost:3210 - that is it. For a production setup, use the official docker-compose with a database (Postgres + pgvector) to enable file RAG and multi-user auth. ## Using local models Set an OpenAI-compatible provider to point at Ollama or vLLM: 1. Go to **Settings** - Language Model - Add provider. 2. Choose OpenAI (compatible), set base URL to `http://localhost:11434/v1` (Ollama). 3. Pick any Ollama model from the dropdown and chat. ## Practical tips - Enable the plugin you need per session to keep context small. - Use the agent store to reuse prompt templates across your team. - Back up your config; the SQLite/Docker volume holds conversations and settings. ## FAQ **Is it free?** Yes - open source (MIT) with optional hosted cloud service. **Does it support RAG?** Yes - with the database deployment, you can chat with documents. **Can multiple users use it?** Yes - with auth enabled and the Postgres setup, it supports multi-user workspaces.