Connect Claude Desktop to Google BigQuery via the Google MCP server (Linux OS)

You can connect Claude Desktop to your Google BigQuery project and get the needed reports or charts by typing plain-language prompts, such as “What is the total revenue generated in the last 7 days?”

The flow is the following:

1. Claude transforms your prompt into an SQL query.
2. The SQL query is sent to the Google Model Context Protocol (MCP) server for execution in Google BigQuery.
3. Google BigQuery returns the query results to the MCP server, which then returns them to Claude.
4. Claude Desktop displays the requested information.

Before proceeding with the configuration, ensure the following:

• You have an active subscription to the AI Datalink module in the Add-on Mart.
• AI Datalink module is deployed.
• Your Google email address has Access to Google BigQuery project.

The configuration steps are described for Linux. If you use Windows, refer to Connect Claude Desktop to Google BigQuery via the Google MCP server (Windows OS). If you use macOS, refer to Google’s official documentation for setup instructions.

1. Download and install Claude Desktop on your computer. Claude Desktop is not officially available for Linux. You can find community-maintained builds through GitHub communities. Alternatively, use the official CLI instead.
   Log in to Claude using the Google email address that has access to your Google BigQuery project.
2. Install Google Cloud CLI to authenticate your Google account, provide credential tokens for the MCP server, and allow Claude Desktop to access your BigQuery project:
   1. Install the prerequisites to make sure your system can securely add and use Google Cloud’s package repository:

      sudo apt-get install apt-transport-https ca-certificates gnupg curl

   2. Add Google’s public key for authentication:

      curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo gpg --dearmor -o /usr/share/keyrings/cloud.google.gpg

   3. Add Google Cloud’s software repository:

      echo "deb [signed-by=/usr/share/keyrings/cloud.google.gpg] https://packages.cloud.google.com/apt cloud-sdk main" \ | sudo tee -a /etc/apt/sources.list.d/google-cloud-sdk.list

   4. Update the packages:

      sudo apt-get update

   5. Install Google Cloud CLI:

      sudo apt-get install google-cloud-cli

3. Authorize access to Google Cloud by running the following command in the terminal:

   gcloud init
   

   You are prompted to select your Google BigQuery project and configure the default region and timezone. Once completed, you’ll see a browser notification confirming successful authorization.

   

4. Set up Application Default Credentials – this provides authentication and authorization for Google MCP Toolbox:

   gcloud auth application-default login

5. Install Google MCP Toolbox:
   1. Download the MCP Toolbox binary:

      curl -O https://storage.googleapis.com/genai-toolbox/v1.2.0/linux/amd64/toolbox

   2. Make it executable:

      chmod +x toolbox

   3. Verify the installation by running

      ./toolbox --version

      If successful, the installed version number appears.

6. Connect Claude Desktop to your Google BigQuery project:
   1. Open Claude Desktop and navigate to Settings.
   2. In the Developer tab, click Edit Config to open the configuration file.

      

   3. In the claude_desktop_config file, add the MCP configuration with the path to the downloaded MCP toolbox and your Google BigQuery project name.

      

7. Save the changes and restart Claude Desktop. You can verify the connection in Settings > Developer.

   

Now, you can type a request in the Claude chat and retrieve the information from your Google BigQuery project.