Bulk address generator
Make a file of test addresses, from 1 to 50,000 records, for one country or many at once. Generation runs in your browser in a background thread, so the page stays responsive even on a phone.
100 addresses · 9 fields · CSV
What is in the file
Each row is one address. Columns use stable snake_case names, so scripts keep working when you add or remove fields. The postal_status column records where each postcode came from:
| Value | Meaning |
|---|---|
| verified | From the same dataset record as the city and region |
| prefix-verified | The routing key or prefix is from the dataset; the rest is synthetic (Ireland, Malta) |
| single | The territory uses one postcode |
| region-prefix | Placeholder that keeps the region’s prefix |
| format-only | Placeholder that only matches the national pattern |
| omitted | Left empty because no dataset links codes to places |
| not-used | The country has no postcodes |
Example
street_line,unit,district,city,region,postal_code,country_code,country,postal_status 9142 River Way #11,,,Spring Creek,TN,38378,US,United States,verified 135 Harbour View,,,London,England,SE15 3DB,GB,United Kingdom,verified Mühlenstraße 31,,,Jungingen,Baden-Württemberg,72417,DE,Germany,verified 福岡市博多区三筑1丁目5-24,,三筑,福岡市博多区,福岡県,812-0887,JP,Japan,verified
Common questions
- Can I get the same file again?
- Yes. Record number n depends only on the seed, its country and n, so the same seed, countries, count and options give a byte-identical file for data version 2026.10.1. A new data version can change the output.
- Why are some rows skipped?
- Territories without civilian addresses (such as Antarctica) produce no rows. The country picker hides them, so this only happens if you add them deliberately.
- How do non-Latin addresses look in Excel?
- CSV files are UTF-8 with an optional byte order mark, which Excel needs to show Japanese, Arabic or Cyrillic text correctly. Add the
*_latincolumns if your system only accepts Latin script. - Is there an API?
- Not yet. Everything runs in the browser from static data files. For automated tests, export a fixture once with a fixed seed and commit it with your tests.
Need people as well as addresses, or SQL output? Try the test data generator.