Mock Data Generator (Fake JSON, CSV, SQL)

Realistic fake data makes it easy to fill a database, demo a UI or test an import without touching real customer records. Describe the columns you need, one name:type per line, and the generator produces up to 1,000 rows as JSON, CSV, TSV, SQL INSERT statements, XML or NDJSON. A seed makes the output repeatable, so a test fixture stays the same every time you regenerate it.

How to use this tool

1. List the fields, one per line, e.g. email:email or age:int(18,90).
2. Set the number of rows (1–1000).
3. Pick the output format and, for SQL or XML, the table or root element name.
4. Keep the seed to get the same data again, or clear it for fresh data.
5. Copy the output.

Formula

name:type(arguments) — int(min,max) · float(min,max,decimals) · date(YYYY-MM-DD,YYYY-MM-DD) · enum(a|b|c) · words(n) · const(text) · rowNumber(start)

Values come from a seeded pseudo-random generator (mulberry32). Within a row the name, email and username belong to the same person, and city, country and country code to the same place. Integers are uniform between min and max inclusive. CSV follows RFC 4180 (quotes doubled, CRLF line ends); SQL doubles single quotes and quotes unusual table or column names.

Example

The fields id:rowNumber and status:const(active) with 3 rows as CSV give exactly four lines: the header id,status, then 1,active, 2,active and 3,active. A value containing a comma or quote, such as He said "hi", ok, is written as "He said ""hi"", ok".