home / mcp / google ads mcp server

Google Ads MCP Server

Enables an AI to query Google Ads data via GAQL.app with a simple Node.js MCP using a token-based stdio interface.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "trueclicks-google-ads-mcp-js": {
      "command": "npx",
      "args": [
        "-y",
        "@trueclicks/google-ads-mcp-js",
        "--token=YOUR_GPT_TOKEN_HERE"
      ]
    }
  }
}

Google Ads MCP (Node.js) lets your AI assistants securely access and query your Google Ads data. It uses GAQL.app as a backend, works on Windows and macOS (Intel and ARM), and requires no Google Cloud setup or OAuth credentials. You configure an MCP client to run the Node.js package and point it at GAQL.app through a simple token mechanism. This setup is designed for a quick, low-friction start so you can begin querying Google Ads within minutes.

How to use

You use an MCP client to connect your AI to the Google Ads MCP server. In practice, you run the MCP as a local stdio server that your AI can communicate with, or you point an MCP-enabled assistant at the GAQL.app backend. The important part is that your AI sends queries through the MCP to Google Ads data without handling OAuth, client IDs, or developer tokens directly.

How to install

winget install nodejs
```

```sh
# On macOS, install Node.js from the official site
# Visit https://nodejs.org and follow the installer

Configuration and running the MCP

You configure the CLI-based MCP by running the tool via npx and supplying your GPT token. The configuration is provided to your AI client through a small JSON setup that your Claude-like product loads. The following snippet shows how to wire the MCP into Claude via a local stdio server.

{
  "mcpServers": {
    "gads": {
      "command": "npx",
      "args": [
        "-y",
        "@trueclicks/google-ads-mcp-js",
        "--token=YOUR_GPT_TOKEN_HERE"
      ]
    }
  }
}

Initial setup steps for Claude (desktop)

Follow these steps to enable Claude to use the Google Ads MCP: open Claude, access Developer settings, and edit the claude_desktop_config.json file. Paste the MCP configuration shown above into the appropriate section so Claude launches the gads MCP server when you start your assistant.

Steps in brief: - Open Claude - Press CTRL + , (Windows/Linux) or Command + , (macOS) - Go to Developer and Edit config - Save the JSON snippet with your GPT token filled in - Exit Claude completely and restart it to apply changes.

Example prompts you can try

List my Google Ads accounts What is the cost for account XYZ in the past 30 days? What are the top 5 setting recommendations for my campaigns?

Notes and considerations

This integration is an unofficial Google Ads MCP. It emphasizes ease of use by avoiding Google Cloud project setup, OAuth credentials, developer tokens, and client IDs. All authentication and query execution is handled behind the scenes by GAQL.app, simplifying setup and reducing configuration friction.

If you encounter issues, ensure Node.js is installed, your GPT token is correctly placed in the configuration, and Claude is restarted after applying settings. Reach out to the provided contact if you need further assistance.

Available tools

GAQL backend

Provides a secure backend that handles authentication and query execution for Google Ads data via GAQL.app.

Claude integration

Allows Claude-like AI assistants to load a local MCP configuration and route Google Ads queries through the MCP to GAQL.app.

Query capability

Supports common Google Ads inquiries such as listing accounts, costs by account and period, and campaign optimization recommendations.