LangBot is an open-source, production-grade platform for building agentic AI bots for instant-messaging apps, connecting large language models to the chat platforms your community or customers already use.
With LangBot, one codebase powers bots across Discord, Telegram, Slack, LINE, QQ, WeChat and WeCom, Lark, DingTalk, KOOK, Matrix and more. Bots can hold multi-turn conversations with tool calling and streaming output, answer from a built-in RAG knowledge base, and plug into agent platforms such as Dify, Coze, n8n and Langflow. Nearly every major model provider is supported – OpenAI, Anthropic, DeepSeek, Gemini, xAI, Moonshot and others, plus local models through Ollama – and a plugin ecosystem with MCP support extends bots with hundreds of community components. Production concerns are covered out of the box: a web management panel (no YAML editing required), access control, rate limiting, sensitive-word filtering, monitoring and multi-pipeline routing for different scenarios. Deployment ranges from a one-line uvx launch or Docker Compose to the hosted LangBot Cloud. The project is Apache-2.0 licensed and under active development, with its latest push in September 2026.
LangBot is free and open source (Apache-2.0) for self-hosting. An official hosted option, LangBot Cloud, is offered for zero-deployment convenience, making the overall model freemium; model API usage is billed by your chosen provider.
LangBot is for community managers, support teams and developers who want capable AI agents living inside chat apps – answering questions, moderating groups and automating workflows – without building the messaging plumbing themselves.