First-time advice: If you haven't isntall OpenClaw yet, I recommend to first proceed to Tencent Cloud OpenClaw to purchase the VPS，you may refer OpenClaw Complete Guide
SPO for one-click completion within a few seconds.

After your instance started, please return to this page, and explore its capabilities.

This article is to explore the method of connecting to WhatsApp after installing OpenClaw in Lighthouse.

Preliminary Preparations

OpenClaw (formerly Clawdbot) natively supports integration with popular messaging platforms such as WhatsApp, Discord, Telegram, and Slack, allowing users to interact with OpenClaw directly through these applications. Before officially starting the WhatsApp integration setup for OpenClaw, please make sure the following requirements are prepared:

You already have one or two WhatsApp accounts (the number depends on whether you want to assign a dedicated account to OpenClaw). Recommendation: preparing a dedicated WhatsApp account for OpenClaw provides a better experience.

Confirm whether OpenClaw has already been deployed on a Lighthouse instance (an overseas region server is recommended). If not yet deployed, it is recommended that:

If you would like to purchase a new Lighthouse server to install and run OpenClaw, you can visit the Tencent Cloud Lighthouse purchase page, or place a discounted order through the Tencent Cloud OpenClaw promotional event. For detailed deployment and configuration steps, please refer to the One-Click Cloud Deployment Guide for OpenClaw (formerly Clawdbot).
If your Lighthouse server using the OpenClaw application template was created before February 11, 2026, it is recommended to follow the tutorial to update the application version: How to Update the OpenClaw Application Version on Your Server.

Step 1. Configure the OpenClaw Model

1.Log in and enter the Console.

2.Locate "Tencent Cloud Lighthouse".

3.Find the Instance you just purchased.

4.Go to the "Pre-installed application" tab.

5.Choose a suitable model/existing API

The Models list only displays models that are available for use in this region:
due to access or service restrictions on some models in specific regions, the system will not display models that are not currently supported in the current region.

If this is your first time using this service, after selecting a suitable model, click "Get API KEY" below to purchase the model's credit limit and obtain the key.

Step 2: Integrating with WhatsApp

1.In the application channel settings, select WhatsApp.

2.Click Authorization Access

3.A confirmation otification will pop up; click "OK".

4.Please wait a moment until a QR code appears.

Open WhatsApp on your phone
go to Settings > Associated device > Associate device, then scan the QR code below.

📢 Important Update for WhatsApp Integration

In Lighthouse, OpenClaw on WhatsApp now uses Private Mode by default, meaning the bot only responds in configured private conversations and will not automatically interact across all chats.

The following image indicates successful deployment.

Search for yourself, converse with yourself—this is your robot.

Now you can chat with the integrated OpenClaw bot in WhatsApp.

More Practical Scenarios

After integrating OpenClaw into Discord, it becomes much more than “an AI you chat with.” It acts as a continuously online personal Agent capable of directly operating cloud servers. Through natural language conversations, we can hand over repetitive daily tasks directly to it.

Scenario 1: Scheduled Reminders (The Simplest AI + cron Workflow)

Simply describe your needs directly in WhatsApp, such as daily weather reminders, meal reminders, or trash disposal reminders at fixed times every day.

OpenClaw will first confirm the necessary information (such as the city), and then create the corresponding cron task.

Throughout the entire process, we never need to directly access the server or write cron expressions ourselves. We simply describe our requirements naturally, as if talking to another person, and the task becomes a long-running automated workflow.

Afterwards, reminders are delivered to the main conversation on schedule, and the Agent only appears “when I need to know.”

Scenario 2: File Organization & Merging (Conversation as the Interface)

When we ask OpenClaw to inspect a directory, read multiple files, and merge their contents, it is not merely “pretending to understand” — it actually enters the server file system and executes the operations directly:

Check the number of files under the specified path
Read the contents of each file one by one
Merge them into a new file and save it as requested
Finally return both the operation result and a structured summary to me

WhatsApp functions more like a remote control console here, while OpenClaw acts as the executor.

What we see is not raw command-line output, but already-organized, human-readable results.

Scenario 3: Generating and Running Code (From “Writing” to “Executing”)

With just a single instruction like “Create a Python file that prints hello world,” OpenClaw can complete the entire workflow:

Create the script file in the specified directory
Write standard Python code into it
Execute the script for verification
Return the execution result

Here we can see that the Agent does not just write code — it also verifies whether the code can actually run successfully. Every reply inside WhatsApp corresponds to an actual task execution performed by the Agent behind the scenes.

Channel	Scan / Click to join
WhatsApp Channel
WeCom (Enterprise WeChat)

Mastering OpenClaw | A Quick Guide to Integrating OpenClaw (Clawdbot) into WhatsApp on the Cloud

Preliminary Preparations

Step 1. Configure the OpenClaw Model

Step 2: Integrating with WhatsApp

More Practical Scenarios

Scenario 1: Scheduled Reminders (The Simplest AI + cron Workflow)

Scenario 2: File Organization & Merging (Conversation as the Interface)

Scenario 3: Generating and Running Code (From “Writing” to “Executing”)

More Tutorials

Welcome to join the discussion!

🚀 Developer Community & Support

1️⃣ OpenClaw Developer Community

2️⃣ Dedicated Support