Power My Analytics gives you two ways to work with your warehoused marketing data programmatically: the PMA API and the PMA MCP Server. This article explains what each one does, who can access it, and how to choose between them.

The short version

Both tools reach the same warehoused, deduplicated, time-zone-normalized data that powers your Data Studio and Google Sheets reports. The difference is how you talk to it:

• The PMA API is for code: it gives your scripts, custom dashboards, and server integrations direct REST access to your data.
• The PMA MCP Server is for conversation: it connects an AI assistant (Claude or ChatGPT) to your Hub so you can ask questions in plain language and get answers from your live data.

PMA API

The PMA API is a REST interface to your warehoused marketing data. You generate a key, authenticate with it, and call endpoints from your own code.

• What it's for: custom dashboards, automated scripts, and server-to-server integrations that need raw, programmatic access to your data.
• Who can use it: all PMA users (see the note on limits below). API access was previously limited to the Custom plan; that restriction has been lifted.
• How to connect: in your Hub, go to Tools > API and click + Create API Key, then authenticate your requests with the header authorization: Bearer YOUR_API_KEY.
• Best for: developers and technical teams who want to pull data into their own systems on their own schedule.

PMA MCP Server

The PMA MCP Server connects Claude, ChatGPT, and other MCP-compatible AI assistants directly to your Hub. Instead of building a report to answer a one-off question, you just ask, and your assistant queries your actual data for you.

• What it's for: natural-language analytics, schema exploration, anomaly detection, and Data Builder dataset management through an AI assistant.
• Who can use it: all PMA users. The MCP Server is in Beta; no early-access request, allow-listing, or admin role is required. Any user with access to your organization can connect.
• How to connect: in your Hub, go to Organizations > MCP, click Connect AI, pick your client, and follow the setup guide. Your token is created automatically through the secure OAuth flow, so there's no manual API key step.
• Supported clients: Claude Desktop, Claude Web, Claude Code, and ChatGPT.
• Best for: marketers and analysts who want fast, ad hoc answers without writing code or building a report.

The MCP Server connects to your data through the PMA API behind the scenes. When you connect an AI assistant, your API token is created for you automatically, so you don't need to generate one first.

Side-by-side comparison

Which one should I use?

• Choose the MCP Server if you want to ask questions and get answers fast, explore your data conversationally, or build quick analyses without code. This is the right starting point for most marketers and analysts.
• Choose the API if you're a developer building a custom dashboard, automating a data pull into your own system, or running a server-to-server integration.
• You can use both. They draw on the same warehoused data, so an answer you explore in an AI assistant can become a query you automate through the API.

A note on cost and limits

Both the MCP server and the PMA API are free to use during their current soft-launch period. For the MCP server, usage-based billing begins August 1, 2026. The PMA API is also in a free soft launch, with usage-based limits and pricing planned for a future date that we'll announce separately. In every case, the specific usage limits and pricing per plan are still being finalized, and we'll give you advance notice before any charges begin.

Related articles

• Power My Analytics MCP Server Guide
• PMA MCP Server Frequently Asked Questions
• Get Started with Power My Analytics API
• How to Generate a New API Key