Skip to main content
The Model Context Protocol (MCP) is an open protocol designed by Anthropic to standardize how applications provide information and context to large language models (LLMs). It defines a consistent, structured way for LLMs to securely query external data through lightweight local servers. The Pagos MCP Server allows you to connect Pagos AI to your preferred AI client (e.g. Claude, ChatGPT), giving you conversational access to payments intelligence through two tools:
  • BIN Database queries - Request standard and enhanced data fields for a given BIN number from the Pagos BIN Database
  • Pagos Data queries - Analyze your business’ harmonized payments data and pull data on transactions events, costs, and conversion

​
Setup and Authentication

You can connect the Pagos MCP Server to your preferred AI chat client. In this guide, we outline the steps for connecting via Claude and ChatGPT.

​
Claude

To configure the Pagos MCP Server via Claude:
1

Connect Your Payment Processors to Pagos

After your business subscribes to Pagos, connect your payment processors using our no-code integrations or the Pagos Data Ingestion API. We’ll ingest payments data from each PSP, then aggregate and harmonize it into a single dataset.
2

Add a Custom Connector

Navigate to the Connectors tab of your Claude account settings, then click Add custom connector.
3

Add the Pagos Connector

Enter Pagos MCP in the Name field, paste https://mcp.pagos.ai/mcp in the Remote MCP server URL field, then click Add.
4

Log In and Allow Permissions

Back in the Connectors tab, click Connect. A new window will open, prompting you to log into your Pagos account. Log in, then click Allow to grant Claude access to your Pagos data.
If you prefer to configure the Pagos MCP Server in Claude code, paste the following into terminal: 
claude mcp add --transport http pagos https://mcp.pagos.ai/mcp

​
ChatGPT

To configure the Pagos MCP Server via ChatGPT:
1

Connect Your Payment Processors to Pagos

After your business subscribes to Pagos, connect your payment processors using our no-code integrations or the Pagos Data Ingestion API. We’ll ingest payments data from each PSP, then aggregate and harmonize it into a single dataset.
2

Navigate to App Settings

Log into ChatGPT, click Apps in the main navigation, then click the Settings icon in the top-right corner.
3

Create a New App

In the Advanced settings, click Create app. Enter Pagos MCP in the Name field, paste https://mcp.pagos.ai/mcp in the MCP Server URL field, then click Create.
4

Log In and Allow Permissions

A new window will open, prompting you to log into your Pagos account. Log in, then click Allow to grant Claude access to your Pagos data.

​
BIN Data Queries

With the Pagos MCP Server, you can use AI agents to query BIN data directly from our BIN Database for a given BIN number. Authenticate with your Pagos account, then use natural language to request all known data points for a given BIN or BIN range, or request specific BIN data. For example, “What issuing bank is associated with BIN 41472024?” or “Is BIN 41472024 a virtual card?”

​
Pagos Data Queries

Pagos Data Queries are available via the Pagos MCP Server for all Insights subscribers.
After you configure the Pagos MCP Server and authorize an AI agent to access your Pagos account, you can query your aggregated payments data conversationally and integrate payment intelligence directly into your AI workflows. Request transaction, refund, chargeback, and cost data across any dimension (e.g. processor, region, card brand) and metric (e.g. count, amount, AOV) within the chat interface; if you can pull the data in a custom Chart in Insights, you can request it via your connected AI client. Keep in mind:
  • The Pagos MCP Server doesn’t share data across Pagos customers; you can only access your own payments data ingested from your connected data sources
  • At this time, you can’t access Benchmarking or Alerts via the Pagos MCP Server.
  • You can provide your AI client with additional context about your business and payment operations to enrich its analysis. Any advice, interpretation, or suggested next steps based on that context comes from your AI client, not Pagos.

​
Troubleshooting

If you need help setting up the Pagos MCP Server, contact [email protected].