EU Eurostat Monitor: Datasets & New Figures (MCP)

The entire Eurostat statistical database — 6,000+ official EU datasets (GDP, inflation, unemployment, energy, migration, trade, demography…) — through one clean API-call format. Find datasets by keyword, fetch filtered datapoints, and get alerted when new figures or revisions drop. Official JSON-stat API, no key, no login. MCP-compatible and agent-ready.

Why this is hard (and worth it)

• Eurostat's own API speaks JSON-stat 2.0: values stored as a flat dictionary keyed by position across a product of dimensions (freq × unit × indicator × geo × time). Nobody — human or LLM — writes that by hand. This actor decodes it into plain rows: time, geo, unit, value.
• Finding the right dataset is its own science: codes like nama_10_gdp, prc_hicp_manr, gov_10dd_edpt1. Search mode scans the full catalogue and gives you codes + titles + freshness.
• Revisions matter: Eurostat rewrites history (GDP gets revised for years back). Monitor mode catches both new observations and revised values — ask any economist how often that has embarrassed a newsletter.

Three modes

search — find the dataset you need

{ "mode": "search", "query": "harmonised inflation monthly", "maxResults": 10 }

→ dataset code, official title, last update, data range, value count.

dataset — datapoints, decoded

{ "mode": "dataset", "datasetCode": "nama_10_gdp", "geo": "DE", "filters": "unit=CP_MEUR&na_item=B1GQ", "timeFrom": "2019", "maxResults": 10 }

→ one row per datapoint: {geo: "DE", time: "2024", value: 4096190, unitLabel: "Current prices, million euro"}, plus the exact API URL and a Data Browser link.

monitor — new figures & revision alerts

{ "mode": "monitor", "watchDatasets": ["nama_10_gdp?unit=CP_MEUR&na_item=B1GQ&geo=DE", "prc_hicp_manr?unit=RCH_A&coicop=CP00&geo=FR"], "maxResults": 30 }

• Run 1 = free baseline across your specs (all geos snapshot at their latest figure).
• Run 2+ emits only: new-value (a newer time period appeared) and value-changed (a revision!) with old → new in info.
• Schedule daily; when Eurostat is quiet, a run costs $0.001 (just the start).

How to find dimension values (unit/na_item/geo/…)

Open the dataset in the Eurostat Data Browser (every result row links to it), pick your filters, and copy the dimension names into filters / watch specs. Popular ones are pre-filled in the examples.

Recipes

1. Agent tool (MCP) — your AI agent answers "German GDP 2024?" or "latest French HICP inflation" with real numbers and the API URL. No hallucinated statistics, ever.
2. Economist bot — schedule monitor with your indicator list; get a dataset row the morning Eurostat publishes (or silently revises) a figure.
3. Research data pull — dataset mode with timeFrom + geo = clean CSV-ready datapoints for any Eurostat table.

Data honesty & limits

• Rows come only from the official Eurostat statistics API at run time; nothing is synthesized. Missing values (denote :, : flags) are omitted — documented in the EU methodology.
• valuesInDataset shows the slice size after your filters so you know what you're sampling from.
• The catalogue snapshot (search mode) is fetched live each run (≈6,000 datasets, ~2 MB text).

Pricing

event	price
Actor start	$0.001
Eurostat datapoint extracted (dataset row or monitor change)	$0.001 / row

Search rows are datasets (metadata). Typical runs: keyword search ×10 ≈ $0.011 · daily monitor, one new figure ≈ $0.002.

MCP / AI-agent usage

https://api.apify.com/v2/acts/YOUR_HANDLE~eu-eurostat-monitor/run-sync-get-dataset-items?token=YOUR_TOKEN

Input = the JSON examples above. Output = rows as documented. Compatible with per-request agentic payments.

Questions / dataset requests: Issues tab on Apify Store.