​
About LangBot

LangBot is an open-source instant messaging bot platform that supports integration with Feishu, DingTalk, QQ, WeChat, Discord, Telegram, and various other messaging platforms. It helps users handle tasks or act as part of a business workflow directly within these platforms.

In combination with the SiliconCloud large model API service, LangBot can integrate with various mainstream open-source large language models. Users can select the most appropriate model based on their specific scenario needs. The following is a complete configuration tutorial — just a few simple steps to get your own intelligent assistant.

​
Configuration Tutorial

​
Step 1: Obtain a SiliconFlow API Key

​
How to Get an API Key

  1. Register and log in to SiliconFlow.

  2. Navigate to the API Key page and generate a new API key.

​
Retrieve Model Information

Visit the Model Marketplace to view available models and detailed parameters.

​
Step 2: Deploy and Configure LangBot

​
Deploy LangBot Using Docker

For more deployment methods, refer to the documentation

Make sure Git and Docker are installed.

git clone https://github.com/langbot-app/LangBot  
cd LangBot
docker compose up -d

If you are using it in mainland China, you can replace the image in docker-compose.yaml with:

docker.langbot.app/langbot-public/rockchin/langbot:latest

​
Access the WebUI

After starting, visit:

http://127.0.0.1:5300

The first run will prompt you to create a configuration file. Follow the instructions to complete initialization.

​
Configure the Chat Model

  1. Log in to the WebUI and navigate to the Model Configuration page.
  2. Add a new model and fill in the following information:
FieldDescription
Model NameChoose the desired model name (available on SiliconFlow’s Model Marketplace)
ProviderSelect SiliconFlow
API KeyUse the key from the API Key page on SiliconFlow
CapabilitiesSelect the appropriate capabilities based on the model description

You can find the model name in the title of the model details page on SiliconFlow.

​
Step 3: Connect to a Platform (Using DingTalk as an Example)

For more platforms like WeCom, Feishu, Discord, Telegram, QQ, WeChat, etc., refer to the LangBot Documentation

​
Create a DingTalk Bot App

  1. Log in to the DingTalk Developer Console

  1. Enter your organization, click App Development → Create App, and fill in the basic information.

  2. Add the Bot capability, complete the basic settings, and publish the app.

​
Configure the Bot

  • Under the Bot tab, fill in the relevant information and publish.
  • In Version Management, configure the version number.
  • In Event Subscriptions, choose Stream Mode (no public callback URL needed).
  • In Credentials & Basic Info, note the following:
    • Client ID
    • Client Secret
    • RobotCode
    • Bot Name

​
Bind in the LangBot Platform

  1. Open the LangBot WebUI and edit the bot.

  1. Bind a pipeline (by default, ChatPipeline is available). Select DingTalk as the platform.

  2. Edit the pipeline, choose Built-in Agent under AI capabilities, and select the previously configured model.

​
Step 4: Use the Bot

  1. In DingTalk, search for the bot name and click to start chatting.

  2. To use the bot in a group chat, go to the group settings, click Add Bot, and search for the bot name to add it.