Skip to content

Connect WhatsApp to Claude on Windows — Without Compiling Anything

In short: Open-source WhatsApp MCP servers ask you to build them, and on Windows that means turning on CGO and installing a C compiler. LMCP is a signed installer: link WhatsApp with a QR code and Claude, ChatGPT or Cursor can read, search and send your chats on your PC.

L
LMCP··5 min read

Average 5.0 out of 5 · Ratings: 2See all reviews →

  • “The no-tokens part is huge. Setting up Graph API and dealing with auth flows is what usually kills these setups for me before I even start.”

    Lisa S.
  • “This is AMAZING! My daily work can all be managed from my key apps now. The OmniFocus integration was the missing piece.”

    Steve H.

To connect WhatsApp to Claude Desktop on Windows, install LMCP and scan one QR code from your phone. LMCP is a local MCP server for Windows: it links to WhatsApp the way WhatsApp Web does, and your AI can then list your chats, read and search messages and send replies. Nothing to compile, no terminal and no config file to edit.

What Your AI Can Do with WhatsApp on Windows

  • List your chats — recent one-to-one and group conversations (whatsapp_list_chats)
  • Read a conversation — the messages of any chat (whatsapp_read_messages)
  • Search messages — find a message by keyword across chats (whatsapp_search_messages)
  • Find a contact — look up who to write to (whatsapp_search_contacts)
  • Send a message or a file — with a preview first and your confirmation (whatsapp_send_message, whatsapp_send_file)
  • Groups and polls — group details, new groups, sending and reading polls (whatsapp_group_info, whatsapp_create_group, whatsapp_send_poll, whatsapp_read_poll)

LMCP executes local tools on your device. Cloud AI providers can receive the messages you ask for; their terms apply.

Why the WhatsApp MCP Servers You Find Need a Build on Windows

Search for a WhatsApp MCP server for Claude Desktop and the results are repositories you build and run yourself. One of them, lharries/whatsapp-mcp, lists Go, Python and uv as prerequisites, and its README has a Windows section: its SQLite driver needs CGO, which is off by default on Windows, so you enable it and install a C compiler (the README suggests MSYS2) before the bridge will run. Then you keep the Go bridge and the Python server running and point Claude Desktop at them in its JSON config.

On WindowsLMCPlharries/whatsapp-mcp
What you installLMCP-Setup.exeGo, Python, uv, a C compiler, the repository
Build stepNoneEnable CGO, run the Go bridge
Linking WhatsAppQR code in the chat or from the trayQR code in the bridge's terminal
Claude Desktop configSet up by the installerEdit the JSON by hand
Next to WhatsAppOutlook, Teams, OneDrive and your files in the same installWhatsApp only

If you want a WhatsApp-only server you can read and modify, the open-source project is a good choice. If you want WhatsApp working in Claude this afternoon, next to the rest of your PC, install LMCP.

How to Install

Download LMCP for Windows and install it:

  1. Run the downloaded LMCP-Setup.exe from your Downloads folder
  2. Follow the installer — no admin rights needed, it installs to your user folder
  3. LMCP starts automatically and appears in your system tray
  4. Restart your AI client (quit Claude Desktop completely and open it again)

Link WhatsApp with a QR Code

  1. Ask your AI: “Connect my WhatsApp.” It runs whatsapp_connect, tells you what will happen and, when you confirm, shows a QR code right in the chat. You can also click the LMCP icon in the system tray, open My apps and click Connect in the WhatsApp row.
  2. On your phone, open WhatsApp → Settings → Linked Devices → Link a Device, and scan the code.
  3. Ask your first question. LMCP shows up in your phone's list of linked devices, like WhatsApp Web.

Example Prompts

  • “Summarize the family group chat from today”
  • “Did the supplier reply about the invoice on WhatsApp?”
  • “Find the address someone sent me last week”
  • “Reply to Ana that I'll be there at 8” — you see the preview before it is sent

Because WhatsApp lives in the same LMCP server as Outlook and Microsoft Teams, your AI can take a detail from a WhatsApp chat and act on it — draft the email, post the update — in one request.

Limitations

  • An unofficial WhatsApp Web link. LMCP links like any WhatsApp Web device. It is not affiliated with WhatsApp or Meta, and WhatsApp may restrict accounts that misuse it: avoid bulk or automated sending.
  • Your phone links it once — by QR code, from WhatsApp's Linked Devices.
  • Personal WhatsApp — this is your own account, not the WhatsApp Business API.
  • Sends show a preview — whatsapp_send_message and whatsapp_send_file send only when you confirm it.

Frequently Asked Questions

Do I need to compile anything or install Go?

No. LMCP-Setup.exe installs everything WhatsApp needs on Windows. There is no Go, Python, CGO or C compiler to set up and no MCP JSON to edit by hand.

How do I link my WhatsApp account?

Ask your AI to connect WhatsApp: it runs whatsapp_connect, which first describes what it will do and then, once you confirm, shows a QR code in the chat. Or click the LMCP icon in the system tray, open My apps and click Connect in the WhatsApp row. Scan the code from your phone: WhatsApp → Settings → Linked Devices → Link a Device.

Can ChatGPT on the web read my WhatsApp on Windows?

Yes. Add LMCP to ChatGPT, Claude.ai or another web AI and turn on Cloud access under Web AI clients in the LMCP tray menu: the WhatsApp tools answer those requests on your PC.

Will it send a message without asking me?

No. whatsapp_send_message and whatsapp_send_file return a preview first; the message goes out only on a second call with confirm=true.

Can I use more than one WhatsApp account?

Yes. Link another one with whatsapp_connect and add_account: true, list them with whatsapp_list_accounts and pick the default with set_default_whatsapp_account.

Related Guides

Ready to try it?

Works with Claude, Cursor, VS Code and any MCP client

Download for Windows
Windows 10+ · 64-bit

✦ llms.txt

Comments

Questions and experiences welcome — comments are reviewed before they appear.