Test data that looks real, a million rows at a time.

Describe the fields, make the rows.

Open the mock data generator and start from Users, Orders, Nested JSON or Hostile input, or add fields of your own: names, emails, addresses, companies, prices, IDs, dates in a range, numbers on a normal curve, a pick from a weighted list, text shaped by a regex, or a template of other fields such as {{first_name}}.{{last_name}}@example.com. Leave a share of a field blank, keep it unique, or put fields inside an object or a list.

Choose how many rows, up to a million, a language and place (17 of them) and the file: JSON, JSON Lines, CSV, TSV, SQL for PostgreSQL, MySQL, SQLite or SQL Server with its CREATE TABLE, XML, YAML or TypeScript. They are made on your device, with a preview of the first rows.

• Open the mock data generator
• UUID v4 vs v7, and every version
• Turn CSV into JSON

From an example you already have.

Paste a JSON row or a list of rows, a JSON Schema or a CSV header, and the fields are made from it: an email column becomes emails, ISO dates become dates, a few repeated values become a pick from them, numbers keep their range. The kinds are judged and can be wrong, so they are shown for you to change.

The same rows, every time.

Rows come from a seed, so the same fields and seed make the same rows on any device: a test can rely on them, and a colleague can make them again from the seed alone.

Nothing leaves your browser.

Everything is worked out on your device: what you paste or open is never sent anywhere, and nothing is kept when you leave. A share link carries what you typed after the #, which browsers never send to a server.

• Read what the site keeps