GeoNames Gazetteer Dumps

Use in your agent

Load GeoNames cities into pandas and Parquet. Checked against the provider's documentation on Nov 1, 2026.

Access

• No account or key. The cities15000 ZIP is a public download.
• The full worldwide and country extracts use the same tab-delimited geoname schema.

Request

curl -O 'https://download.geonames.org/export/dump/cities15000.zip'

Format

ZIP containing headerless, tab-delimited UTF-8 text. Gazetteer columns cover ID, names, latitude, longitude, feature class and code, country and administrative codes, population, elevation, DEM, timezone and modification date.

Load into pandas and Parquet

Needs Python 3 with requests, pandas and pyarrow.

"""GeoNames cities with population over 15,000 or capitals -> pandas -> Parquet."""
import io
import zipfile
import requests
import pandas as pd

URL = "https://download.geonames.org/export/dump/cities15000.zip"
COLUMNS = [
    "geonameid", "name", "asciiname", "alternatenames", "latitude", "longitude",
    "feature_class", "feature_code", "country_code", "cc2", "admin1_code",
    "admin2_code", "admin3_code", "admin4_code", "population", "elevation",
    "dem", "timezone", "modification_date"
]
r = requests.get(URL, headers={"User-Agent": "SignalLabCheck/1.0"}, timeout=60)
r.raise_for_status()
with zipfile.ZipFile(io.BytesIO(r.content)) as archive:
    df = pd.read_csv(archive.open("cities15000.txt"), sep="\t", names=COLUMNS,
                     header=None, low_memory=False)
df["modification_date"] = pd.to_datetime(df["modification_date"])
df.to_parquet("geonames_cities15000.parquet", index=False)
print(len(df), "places", df["country_code"].nunique(), "country codes",
      "latest modification", df["modification_date"].max().date().isoformat())

Run by us on 2026-11-01 (python 3.13.5 | pandas 3.1.0 | pyarrow 26.0.0 | requests 2.34.2). Result: 34195 places 243 country codes latest modification 2026-10-31

Limits

• Download rate limit: Not stated.
• The checked allCountries ZIP was 403 MB compressed; cities15000 was 3.2 MB compressed.
• Dated change files describe the previous day's modifications and deletions.

Quirks

• The text files have no header row; supply the readme's column order explicitly.
• Use alternateNamesV2.zip when language codes and validity dates are needed; the older alternateNames.zip is marked obsolete.
• The alternatenames column in the main geoname table is a shortened convenience field and does not replace the language-aware alternate-name table.

Terms and attribution

• The data is licensed under Creative Commons Attribution 4.0 and commercial use is allowed.
• Credit GeoNames with a link or another reference.
• The provider supplies the data as is without accuracy, timeliness or completeness warranties.

Sources

• https://download.geonames.org/export/dump/readme.txt: “This work is licensed under a Creative Commons Attribution 4.0 License,”
• https://download.geonames.org/export/dump/readme.txt: “allCountries.zip : all countries combined in one file, see 'geoname' table for columns”
• https://www.geonames.org/export/: “GeoNames data is free, the data is available without costs.”
• https://www.geonames.org/export/: “You should give credit to GeoNames when using data or web services with a link or another reference to GeoNames.”