How does offline geocoding work without an API?

The tool ships with an embedded gazetteer of major world cities and country centroids. For each row it normalizes the city name, looks it up in the city table, and if there is no city match it falls back to the centroid of the country. No network request is ever made.

How accurate are the coordinates?

This is a city and country level geocoder, not a street level one. A matched city returns that city's real coordinates; a country-only match returns the country's approximate geographic center. It is built for mapping and clustering at a regional scale, not for precise addresses.

What does the geo_match column mean?

It records how each row was resolved. city means the city name was found directly, country means only the country centroid was available, and none means neither the city nor the country could be matched in the embedded table.

Why is privacy a selling point here?

Because nothing is uploaded. Online geocoders send every address you submit to a third-party server. This tool keeps the entire address list in your browser, which matters for customer lists, patient records, or any data you cannot legally transmit.

Why didn't a city match?

The embedded table covers major metros and capitals, so smaller towns may not be present and will fall back to the country centroid or report none. Adding a country column improves matching and disambiguates cities that share a name across countries.

CSV Offline Geocoder (City / Country Level)

Mapping a spreadsheet of locations usually means sending every address to an online geocoding API — which is a privacy problem for customer, patient, or supplier lists. This tool instead carries an embedded gazetteer of major cities and country centroids, so it adds coordinates to your CSV entirely in your browser with no API call.

How it works

For each data row the geocoder runs a small matching pipeline:

Normalize the city value (lowercase, collapse whitespace) and, if present, resolve the country column to an ISO country code, accepting names and common aliases like UK or USA.
City lookup. It searches the embedded city table, preferring a match in the stated country so that cities sharing a name are disambiguated.
Country fallback. If no city matches, it returns the centroid (approximate geographic center) of the country. If even that fails, the row is marked none.

It appends three columns — lat, lon, and geo_match — and produces a valid RFC 4180 CSV you can copy out.

Example

A row Acme Ltd, London, United Kingdom resolves to 51.5074, -0.1278 with geo_match = city. A row whose city is unknown but whose country is Armenia falls back to the Armenia centroid with geo_match = country. A row like Atlantis, Nowhere yields empty coordinates and geo_match = none.

Notes

Because the dataset is embedded and compact, this is a regional-scale geocoder: ideal for clustering customers by metro, plotting a country breakdown, or sanity-checking a list — not for routing to a doorstep. The big advantage is that your address data never leaves the browser, making it safe for sensitive lists that an online API would expose.