How to Connect Eurostat to Gemini via MCP

💡 No install? Use cloud.anythingmcp.com directly. Sign in, click Connectors → Eurostat, install it with no key, mint an MCP API key and you're done. No Docker, no git clone, no local server.

Eurostat + Gemini

Official statistics of the European Union from Eurostat, no API key: read any of about 8,000 datasets (GDP, inflation, unemployment, population, energy, environment, regional NUTS data) filtered by country, period and dimension, inspect dimensions, codes and availability through SDMX, look up code lists, follow the daily update feed and query detailed Comext trade data.

Prerequisites

See the full setup instructions baked into the connector (visible in the in-app store when you select the connector). The required environment variables for this connector are:

(none: public data, no API key)

Step 1 — Set up the connector

Public data, no key: setup takes under a minute. The connector reads the public dissemination APIs of Eurostat, the statistical office of the European Union (https://ec.europa.eu/eurostat/api/dissemination). They are free and need no account, API key or sign-in.

1. In the connector store, pick Eurostat and install it. There is nothing to paste and no Authorize with Provider step.
2. Check it with eurostat_get_dataset_info and the dataset nama_10_gdp: it shows the title of annual GDP, the number of observations, the oldest and latest period and the last update.

Reuse: the data may be reused under the Eurostat copyright notice. Cite "Source: Eurostat" with the dataset code.

Finding a dataset: there is no keyword search, so the AI works with dataset codes. It knows the common ones, such as nama_10_gdp (annual GDP), namq_10_gdp (quarterly GDP), prc_hicp_manr (HICP inflation, annual rate), une_rt_m (monthly unemployment), demo_pjan (population on 1 January), gov_10dd_edpt1 (government deficit and debt), nrg_pc_204 (household electricity prices), env_air_gge (greenhouse gas emissions) and nama_10r_2gdp (regional GDP by NUTS 2). Any other code appears in the URL and title of the dataset in the Eurostat Data Browser (https://ec.europa.eu/eurostat/databrowser). eurostat_list_recent_updates lists the datasets added, deleted or updated in the last 7 days, and eurostat_list_comext_datasets lists the Comext and PRODCOM datasets.

What it reads:

• About 8,000 datasets on the economy, prices, the labour market, population, health, education, energy, environment, agriculture, transport and regions (NUTS 1 to 3).
• Dataset metadata: title, number of observations, oldest and latest period, last data and structure update, planned next update, source institutions and the methodology (ESMS) page.
• The dimensions of a dataset in key order with every code it uses and its label, the codes that actually have data together with the available periods, the SDMX data structure and whole code lists such as GEO, UNIT, NA_ITEM or NACE_R2.
• Values filtered by country or region (geo, or geoLevel for every country, NUTS 1, 2 or 3 region or city at once), period, unit, freq and any other dimension through filters, one value per dimension.
• Series by SDMX key, when a dimension other than geo, time, unit or freq needs several values, for example A.CP_MEUR.B1GQ+P3.DE+FR.
• Detailed EU trade in goods from Comext by reporter, partner, product (HS chapters down to CN8), flow and indicator, and PRODCOM industrial production.
• The official update feed of the last 7 days, rebuilt at 11:00 and 23:00 CET.

Data formats: values come as JSON-stat 2.0. id lists the dimensions in order and size the number of codes returned for each; dimension.<id>.category maps each code to its position and label; value holds the cells keyed by their flat index, and missing cells are left out. status flags cells, for example p provisional, e estimated, b break in series, c confidential. The AI reads the unit label before comparing figures (million euro, percent or index). Periods look like 2024, 2024-Q3, 2024-09 or 2024-09-30. Eurostat publishes with a delay, so the AI quotes the latest period returned, not today's date. Labels are in English by default, or in French or German with lang.

Time filters: lastTimePeriod 1 returns the latest value; sinceTimePeriod and untilTimePeriod give a range and are the only two time parameters that can be combined. In Comext, lastTimePeriod also counts empty future months, so the AI prefers sinceTimePeriod there.

No paging: a request returns every matching cell at once. That is why the AI filters by country, unit and period to keep answers small.

Read-only: every tool only reads, so there is nothing to confirm and no tool is switched off at install.

Limits: requests estimated above 500,000 cells are not answered at once, and above 5,000,000 cells they are refused with 413 "EXTRACTION_TOO_BIG". An answer with warning.status 413 "ASYNCHRONOUS_RESPONSE" means Eurostat is still preparing the extract: the AI narrows the filters or asks again a little later. Whole code lists are big (GEO has more than 3,000 codes), so for one dataset the AI uses eurostat_get_dataset_codes. The Comext product dimension has tens of thousands of codes, so the AI always sets reporter, product, flow and indicators.

Common errors: 400 names the problem in error[0].label, for example a filter on a dimension the dataset does not have, or geo and geoLevel sent together. 404 "is not available for dissemination" means an unknown or removed dataset code. An empty value means there is no data for that combination: the AI checks the codes with eurostat_get_dataset_availability.

The connector does not download the bulk TSV or SDMX files, does not read the full table of contents and does not use the asynchronous extraction API.

Step 2 — Install the adapter

curl -fsSL https://raw.githubusercontent.com/HelpCode-ai/anythingmcp/main/docker-compose.quickstart.yml -o docker-compose.yml
printf 'JWT_SECRET=%s\nENCRYPTION_KEY=%s\n' "$(openssl rand -hex 32)" "$(openssl rand -hex 32)" > .env
docker compose up -d

Step 3 — Add the connector in Gemini

Your server URL: in AnythingMCP, open MCP Servers → the server this connector is on, and copy its URL (https://cloud.anythingmcp.com/mcp/…). Use it wherever this guide shows YOUR_SERVER_ID.

Gemini CLI reads MCP servers from ~/.gemini/settings.json (on Windows %USERPROFILE%\.gemini\settings.json). Add:

{
  "mcpServers": {
    "anythingmcp": {
      "httpUrl": "https://cloud.anythingmcp.com/mcp/YOUR_SERVER_ID",
      "headers": { "X-API-Key": "YOUR_MCP_API_KEY" }
    }
  }
}

1. Create an MCP API key in AnythingMCP under MCP Servers → your server → API keys, and paste it in place of YOUR_MCP_API_KEY.
2. Save the file and restart gemini.
3. Run /mcp inside Gemini CLI (or gemini mcp list in your shell): anythingmcp should show as connected.
4. Start prompting: Gemini CLI can now call the Eurostat tools.

Available tools

Example prompts

• "How did real GDP grow in Germany, France and Italy over the last five years?"
• "Show the latest annual HICP inflation rate for every euro area country, highest first."
• "How has the monthly unemployment rate in Spain changed since January 2020?"
• "Which NUTS 2 regions of Poland have the highest GDP, and how has it changed since 2015?"
• "How much did Germany export to the United States in vehicles (HS chapter 87) each year since 2022, in euros?"

FAQ

Where does Gemini CLI look for MCP servers? In ~/.gemini/settings.json for your user, or in .gemini/settings.json inside a project. gemini mcp list shows each server and whether it connects.

Next steps

• Connector landing page
• Connect to Claude
• Connect to ChatGPT
• Connect to Cursor
• Connect to Meta Muse
• Connect to Microsoft Copilot