Anthropic Economic Index MCP Server

What is Anthropic Economic Index MCP Server?

Anthropic Economic Index MCP Server is Anthropic's official, free endpoint (https://econ-index.mcp.claude.com/mcp) that lets AI assistants explore the Anthropic Economic Index, Anthropic's public data on how people around the world use Claude, by country, US state, occupation and task.

The Anthropic Economic Index is a research initiative on AI's effects on the economy. It matches aggregated Claude conversations to job tasks and publishes the results as reports and open datasets under a CC BY 4.0 license. The MCP server puts that dataset behind a conversational interface, so researchers, journalists, policy analysts and curious users can ask "Which occupations use AI the most?" and get figures straight from the source. Anthropic lists it in the Claude connector directory as an Anthropic-built connector.

How to use Anthropic Economic Index MCP Server?

No account, API key or sign-in is needed for the server itself. Add it by URL:

https://econ-index.mcp.claude.com/mcp

Claude Code

claude mcp add --transport http economic-index https://econ-index.mcp.claude.com/mcp

Claude Desktop and claude.ai

The connector is listed in the Claude directory as "Anthropic Economic Index". Open the connectors menu in claude.ai settings, find it in the directory and enable it; after that it works in any conversation. You can also add it as a custom connector with the URL above.

Cursor

{
  "mcpServers": {
    "economic-index": {
      "url": "https://econ-index.mcp.claude.com/mcp"
    }
  }
}

VS Code

{
  "servers": {
    "economic-index": {
      "type": "http",
      "url": "https://econ-index.mcp.claude.com/mcp"
    }
  }
}

Key features of Anthropic Economic Index MCP Server

• Official server published by Anthropic, with data licensed CC BY 4.0.
• Free and open: no authentication required.
• Geographic coverage of more than 120 countries plus all 50 US states and Washington, D.C., at the time of writing.
• Occupation and task data: around 22 job categories and several hundred published occupations, with task descriptions drawn from O*NET.
• Usage splits such as augmentation versus automation, and work versus personal versus coursework use.
• Built-in methodology notes and definitions, so the assistant can present the numbers with the right caveats.
• Tool outputs designed for visualization, such as ISO country codes for building a world map.

Anthropic Economic Index MCP Server tools

The server exposes nine tools, all prefixed econ_index_:

• Start here: econ_index_get_dataset_overview returns coverage, the latest data period, license, definitions and the global headline split. It is meant to be called once at the start of an exploration.
• Global view: econ_index_get_global_usage returns the worldwide profile, and econ_index_list_top_work_tasks returns the top 50 work tasks people use Claude for.
• Countries: econ_index_list_countries ranks every country with a published Anthropic Usage Index, and econ_index_get_usage_by_country returns one country's full profile.
• Subregions: econ_index_list_subregions ranks a country's subregions (US states by default), and econ_index_get_usage_by_subregion profiles a single state, province or prefecture.
• Occupations and comparisons: econ_index_get_occupation_usage ranks job categories and drills into occupations, while econ_index_compare_regions compares two to eight regions from the same cohort.

Use cases of Anthropic Economic Index MCP Server

• Occupational research: "Which occupations use AI the most, and what tasks are they using it for?"
• Regional analysis: "What are the most common ways people in Colorado use Claude?"
• Country comparisons: "Compare Claude usage in Japan, Germany and Brazil."
• Journalism: check a claim about AI adoption against Anthropic's published figures and cite the source.
• Teaching and policy work: explain the difference between augmentation and automation using real usage splits.

Troubleshooting Anthropic Economic Index MCP Server

• Numbers look like a trend but are not: the server serves a snapshot of the latest data period. It has no trend series, so it cannot show shares rising or falling over time.
• Index values do not line up: the Anthropic Usage Index is normalized within one cohort. Compare countries with countries, or subregions of one country with each other, never across cohorts.
• Missing values: a null means the value is not published for that cell, not necessarily that it is zero.
• Exact profession not found: the dataset covers occupations with published usage, not the full taxonomy; use the closest published occupation or category.