Does the generated data resemble real people?

The values look plausible, like a name and email pair, but they are assembled from small fixed word lists with a random-number generator, so nothing here corresponds to an actual person or account.

What field types are supported?

Names, email, username, phone, UUID, integers, decimals, booleans, date and datetime, company, city, country, street, URL, IPv4, word, sentence and color are all available from the type dropdown per field.

What is the maximum number of rows I can generate?

Row count is capped at 5000 per generation; requesting more just clamps to that limit, which keeps generation fast since everything runs synchronously in the browser tab.

Can I export the rows as SQL insert statements?

Yes, choosing SQL as the format produces one INSERT INTO statement per row using the table name you provide, with values quoted and escaped for that table.

Fake Data Generator

Generate realistic fake data from a field schema and export as JSON, CSV or SQL, locally.

Local, Runs entirely in your browser, your data never leaves your device.

Loading tool…

Generate placeholder rows without touching real records

This tool generates rows of made-up data from a schema you define: name each field, pick a type such as email, UUID, boolean or city, and it produces up to five thousand rows in JSON, CSV or SQL insert format. It's meant for seeding a test database, filling a UI mockup, or sharing a sample export with a teammate without reusing anyone's real customer data. Every value is generated with plain random logic running in your browser tab.

How to use the Fake Data Generator

Rename the default fields or click + Add to create new ones for your schema.
Pick a type for each field from the dropdown, such as fullName, email, uuid or ipv4.
Set the row count, up to 5000, and choose JSON, CSV or SQL as the output format.
For SQL output, enter a table name, then click Generate to produce fresh rows.
Copy the output or use the send-to menu to pass it into another tool.

Names, companies, cities and words are drawn from small fixed lists rather than a large corpus, so with high row counts you will see repeats; this is expected for a lightweight, dependency-free generator and does not affect its usefulness for structural testing.

SQL output wraps text values in single quotes and doubles any embedded quote, and treats booleans as 1 or 0, but it does not know your target database's column types, so review the generated INSERT statements before running them against a real schema.

Frequently asked questions

Does the generated data resemble real people?: The values look plausible, like a name and email pair, but they are assembled from small fixed word lists with a random-number generator, so nothing here corresponds to an actual person or account.
What field types are supported?: Names, email, username, phone, UUID, integers, decimals, booleans, date and datetime, company, city, country, street, URL, IPv4, word, sentence and color are all available from the type dropdown per field.
What is the maximum number of rows I can generate?: Row count is capped at 5000 per generation; requesting more just clamps to that limit, which keeps generation fast since everything runs synchronously in the browser tab.
Can I export the rows as SQL insert statements?: Yes, choosing SQL as the format produces one INSERT INTO statement per row using the table name you provide, with values quoted and escaped for that table.