Test Data Generator
Choose the fields, the number of rows and a format, and get realistic made-up records for testing software and spreadsheets, with nothing real in them.
Made-up data only, made on your device. It never includes card numbers, bank details, Social Security or other ID numbers, passport numbers or tax IDs.
Realistic test data, with nothing real in it
Testing a sign-up form, a spreadsheet formula or a new database table goes better with data that looks real: names, addresses, dates and numbers of the right shape. Real customer data doesn’t belong in a test, and typing a hundred made-up rows by hand takes an afternoon. This generator makes up to 1,000 rows in a moment, as CSV, JSON or SQL.
How to use it
- Tick the fields you need. Columns come out in the order they’re listed.
- Choose how many rows, and whether addresses and phone numbers are from the US, the UK or both.
- If you chose numbers or dates, set their range.
- Pick CSV, JSON or SQL. For SQL, type the table name and choose your database.
- Copy the output, or download it as a file.
What the options do
- Seed: the same seed with the same settings gives exactly the same rows, so a test can be repeated. Leave it empty for a random one; the seed used is shown above the output.
- CSV: a header row, then one line per record, with commas and quotes inside values handled the standard way.
- JSON: a list of objects, with numbers and true or false values kept as their own types.
- SQL: one INSERT statement per row. Table and column names are quoted for your database, and quotes inside text are escaped, so a name like O’Neill can’t break the statement.
What’s in the data
- Emails use only example.com, example.org and example.net, domains reserved for examples, so a test can’t email a real person.
- Phone numbers come only from the ranges set aside for fiction: 555-0100 to 555-0199 in the US, and 07700 900000 to 07700 900999 in the UK.
- Street names are invented; cities are real, and postal codes are made up to look right for the country.
- Never generated: card numbers, bank details, Social Security or other national ID numbers, passport numbers, tax IDs, or anything resembling a real document.
Good to know
- Emails, usernames and UUIDs don’t repeat within a set. Names, phone numbers and addresses can, as in real life.
- The UUIDs come from the seeded generator: fine for test rows, not for real records or anything secret.
- Cleaning up a real list instead? Remove Duplicate Lines finds the repeats.
Frequently asked questions
Can I get exactly the same data again?
Yes. Every set is made from a seed number, shown above the output. Type that number in the seed box, keep the same settings, and you get the same rows, which helps when a test has to be repeated or shared with a colleague.
Is any of it real?
The cities and countries are real, and the names are common first and last names put together at random, so any match with a real person is chance. Street names, postal codes and companies are made up, emails use example domains only, and phone numbers come from the ranges set aside for fiction.
Why doesn’t it make card numbers or ID numbers?
Made-up numbers in those formats can match real ones, and payment details and identity documents don’t belong in test data. If you’re testing payments, use the test card numbers your payment provider publishes for that.
Which databases does the SQL work with?
PostgreSQL and SQLite, MySQL and MariaDB, and SQL Server, each with its own way of quoting names. There’s one INSERT statement per row; dates are written as text such as 2026-10-20. Create the table with matching columns before you run it.