Copying real customer records into a development database or a sales demo is one of the most common ways personal data leaks, and under privacy laws such as the GDPR that copy needs a legal basis a test environment rarely has. Fake data solves the problem: rows that look like real people, companies and addresses, so layouts, validation and imports behave realistically, without belonging to anyone.
Pick the fields you need, set the number of rows, and generate a JSON array or a CSV file. The data is produced in your browser by Faker, a widely used open-source library, and nothing you generate is sent to or stored on a server.
What fake data is, and why it is safe to use
Fake data, also called mock or dummy data, is generated from lists of plausible first names, surnames, street names, cities and company patterns, combined at random. Each row is internally consistent where it matters for testing: the email address is built from the same first and last name as the Name column, so a record for Laura Jensen gets an address such as laura.jensen@example.com. In Arabic mode the names are in Arabic script and the email user names are transliterated into Latin letters.
All generated email addresses use example.com, a domain reserved for documentation and examples by RFC 2606, so test messages sent to them can never reach a real person. Names and phone numbers are random combinations; any resemblance to a real person or a working number is coincidence, so do not call or text them. The data is fictional and intended only for testing, demos and design work, never for impersonation, fake accounts or misleading anyone.
Where mock data helps
Seeding development and staging databases
Import a CSV of a few hundred fake customers to exercise lists, search, pagination and sorting. Because the data is safe to share, it can be committed to a repository or handed to a contractor.
Automated tests with stable fixtures
Set a seed and the generator returns exactly the same rows every time with the same settings. Tests that compare output against a snapshot stay deterministic, and anyone on the team can regenerate the fixture.
Design mockups and product demos
Real-looking names, job titles and companies make a dashboard screenshot or a sales demo far more convincing than John Doe repeated twenty times.
Testing right-to-left and Arabic interfaces
Switch Language to Arabic to fill a form, table or PDF template with Arabic names and places, a quick way to catch alignment and font issues in bilingual apps.
Fields, seeds and output formats
Thirteen fields are available: Name, Email, Phone, Address (a street address), City, Country, Company, Job title, Date, UUID, Number, Boolean and Sentence. Dates are written as YYYY-MM-DD and fall within the five years before 1 January 2025; Number is a whole number from 1 to 10,000; UUID is a random version 4 identifier; Sentence is Latin-style placeholder text in English mode. Name, Email, Phone, City and Company are selected by default, and the keys in the output are the lowercase field names, such as name, email and job, in a fixed order.
A seed is any whole number. With the same seed, language, fields and row count, the output is identical on every run; leave the box empty for fresh random data each time. JSON output is a pretty-printed array of objects, ready for a mock API or a fixture file. CSV output follows RFC 4180: a header row, commas between fields, quotes around values that contain commas or quotes, and CRLF line endings, which Excel, Google Sheets and database import tools all accept. Switching between JSON and CSV reformats the rows you already have rather than generating new ones.
Como usar esta ferramenta
- Under Fields, switch on the columns you want and switch off the rest.
- Enter the number of rows, from 1 to 1,000, and optionally a seed for reproducible output.
- Choose English or Arabic under Language, and JSON or CSV under Format.
- Press Generate data and review the rows in the output panel.
- Copy the result or download it as fake-data.json or fake-data.csv.
Limitations
- Fields are generated independently, so a row's address, city and country do not belong together, and phone numbers use North American style formats, sometimes with an extension such as x123, whatever the country or language.
- You cannot rename fields, add custom columns, set value ranges or create related tables with foreign keys.
- Output is limited to 1,000 rows per run, in JSON or CSV only; there is no SQL INSERT or Excel export.
- Seeded results depend on the Faker version, so they can change if the library is updated, and other Faker-based tools may give different rows for the same seed.