LangBot is an open-source instant messaging AI bot platform, with the goal of connecting AI applications to the chat scenarios where real users are. It supports more than 30 mainstream instant messaging platforms, including QQ, WeChat (personal accounts and official accounts), WeCom, Feishu, DingTalk, Telegram, Discord, Slack, LINE, KOOK, Mattermost, and continuously adapts to new platforms.
Main capabilities
- Multi-platform access: one codebase for multi-platform deployment, covering mainstream instant messaging platforms at home and abroad.
- Model and platform integration: built-in model channel management, compatible with the OpenAI format, supports mainstream models such as Claude, Gemini, and Ollama, and provides multi-Provider load balancing and automatic fallback; natively supports LLM application platforms such as Dify, Coze, FastGPT, n8n, DeerFlow, Alibaba Cloud Bailian, and LangFlow, and can connect to application types such as chat assistants, Agents, and workflows.
- Conversation scenario features: built-in capabilities such as rate limiting, access control, and context management, can answer questions, process files, welcome new members, and trigger Agent processing when a message is received, someone joins a group, or a friend request is sent.
- Extension mechanism: provides a plugin system and supports developing custom features through the Python SDK; the extension marketplace already has more than 200 extensions available for one-click installation, covering plugins, MCP services, and skills (Skill), and covering scenarios such as scheduled tasks, image generation, knowledge base retrieval, and web search. Versions that pass review will be marked as certified plugins.
- Sandbox: provides an isolated working environment for Agents, used to run code and process files, and users can decide which tools they can use.
- Knowledge base and RAG: built-in RAG engine, supports multiple document formats, and is deeply adapted to external knowledge base services such as Dify and RAGFlow.
- Observability and auditing: full-link log tracing, conversation records can be audited, and performance metrics such as running status, Token usage, and latency can be monitored in real time.
- Multi-tenancy and team collaboration: the same Agent can be used for multiple bots and group chats; the workspace supports inviting members, and each member joins with their own account to jointly manage bots, knowledge bases, and extensions; roles and permissions can be assigned to members; message records, model call records, and user feedback can be exported as CSV.
How to use How to use The simplest way is to use Docker for one-click deployment, which can be started with just one docker compose command; you can also use LangBot Cloud to create an instance online, without deploying a server yourself. The official team also provides prompts that can be copied to a local Agent to assist with deployment, connecting chat platforms, and daily management. For detailed tutorials, please refer to the official documentation.
Applicable scenarios It can be used for scenarios such as customer service bots, internal knowledge assistants, and automated workflows, and is suitable for individual developers and enterprise teams that want to bring AI capabilities to end users. Enterprises can use it commercially for free under Apache 2.0, or choose an enterprise support subscription to obtain limited official support, contracts, and invoices.
Limitations and notes The core features of LangBot are completely free and open source (Apache 2.0 license), and self-deployment is not subject to feature restrictions; the LangBot Cloud hosting service is paid monthly. For specific prices, platform adaptation details, and the latest features, please refer to the official documentation.
