How to Generate Dummy Data for Testing (Safely and Free)

What the Dummy Data Generator Gives You

• 19 field types across Person, Business, Address, Tech, Date, Finance and Text, such as full name, email, phone, company, postcode, UUID, IP address, date, currency amount and credit card number.
• English (US) or English (UK) formats for names, phone numbers, addresses and postcodes.
• From 1 to 500 rows per run.
• JSON, CSV or plain text output, which you copy with one click.

Values come from the open-source Faker library and are generated in your browser. A JSON row looks like this:

[
  {
    "Full Name": "Nikita Crist",
    "Email": "[email protected]",
    "City": "Mooreford"
  }
]

Email addresses always use example.com, example.net or example.org. Phone numbers use the national format without extensions, UK postcodes follow the UK postcode format, and card numbers pass the Luhn checksum. All of them are random, so most numbers and postcodes don't exist. The tool page lists the details for each field.

Edge Cases to Add by Hand

Random data is realistic but rarely hits the boundaries where bugs hide. Every selected field always gets a value, dates fall within the past year and amounts stay between 10 and 10,000. Add a few rows by hand to cover:

• Empty and missing values: blank names, a missing email, a null amount.
• Maximum lengths: a 255-character name, or whatever your database column allows, plus one character more.
• Unusual characters: apostrophes and hyphens (some generated surnames have them), accents such as Zoë or José, and emoji.
• Boundary dates: 29 February, future dates, very old dates and the first and last day of a month or year.
• Invalid input: malformed emails and postcodes, so your validation is tested as well as your happy path.
• Deliberate duplicates: repeated emails or IDs, to check unique constraints.

To check validation patterns against these values, use the Regex Tester.

Repeatable Data and Larger Datasets

The online generator has no seed option, so every run produces new values. If a test needs the same data every time, save the output you test with and commit it alongside your tests.

For more than 500 rows, or for data generated as part of a test suite, use the Faker library in code. Setting a seed makes the output repeatable:

import { faker } from "@faker-js/faker/locale/en_GB";

faker.seed(42); // same seed, same data on every run

const users = Array.from({ length: 1000 }, () => ({
  name: faker.person.fullName(),
  email: faker.internet.exampleEmail(), // example.com, .net or .org
  city: faker.location.city(),
}));

With the same seed and the same Faker version, this produces the same 1,000 users on every run.

Using Fake Data Safely

Generated names, phone numbers and addresses can match real people or businesses by chance, so never send real messages to generated contact details. Use a test inbox or a sandbox service when your tests send emails or texts.

Using generated records instead of copies of customer data can reduce the personal data held in development and test systems. It does not make a project compliant with data protection law on its own: that depends on how you build and run your systems.

Conclusion

Generate realistic rows for the bulk of your test data, add the edge cases by hand, and save the dataset you rely on. Keep generated contact details away from real inboxes and phones.