LangBot

LangBot is an open-source platform for building instant-messaging bots. It connects to Feishu, DingTalk, WeChat, QQ, Telegram, Discord, Slack and other platforms and supports knowledge bases, agents and MCP. Docs: docs.langbot.app, source: GitHub.

Prerequisites

• An API key: open the console, go to API Keys and create a dedicated key (for example, named LangBot).
• A running LangBot instance — see the official deployment docs.

Configuration

1. Add a chat model

1. In the LangBot web UI, open Models and add a new model.

2. Choose an OpenAI-compatible requester (LangBot also lists a dedicated New API requester for this kind of gateway) and fill in:

   Field	Value
   Model name	a model ID from the Model Square, e.g. gpt-4o-mini
   Base URL	https://rutoken.ai/v1
   API Key	sk-xxxx

3. Save the model.

2. Use it in a pipeline

Open Pipelines, select the model you just added for the pipeline's AI settings and save. Bind the pipeline to a bot — see Deploying Bots for each messaging platform.

3. Optional: knowledge base embeddings

RuToken also provides embedding models. Add an embedding model in LangBot with the same Base URL and key (for example text-embedding-3-small), then select it when creating a knowledge base.

Verify

Use Debug chat in the pipeline, or message a bot bound to the pipeline. A reply means the model works; the request appears in the console under Usage Logs.

Troubleshooting