Fake JSON Data Generator

How to generate fake JSON data

1. Add a field for each key you want in the output, and give it a name such as firstName, email or id.
2. Pick a type from the dropdown. Some types show extra options, a min and max for Integer and Float, a date range for Date, or a comma-separated list for Enum.
3. Set how many records you need, from 1 up to 10,000 per batch.
4. Click Generate JSON. Every record gets its own freshly-randomized value for each field.
5. Copy or download the result as a .json file and drop it straight into your project.

Why generate fake structured data?

Plain lorem ipsum text is fine for a paragraph, but it falls apart the moment you need an array of objects that looks like real records. A username field full of repeated placeholder text, or an API response where every row is identical, makes it hard to notice bugs that only show up with varied, realistic-looking values, things like a name that is much longer than average, a price with cents, or a date near the edge of a range. This tool fills that gap: it is a lorem ipsum generator for structured data, built for developers and QA testers who need believable fake records rather than filler sentences.

The most common use is seeding a local database during development. Instead of typing rows by hand or writing a one-off script with hardcoded values, define the shape you need once and generate as many rows as your table needs, from a handful for a quick check to a few thousand for a rough load test.

It is just as useful for mocking an API response before the real backend exists. Front-end work can start immediately against a realistic-looking payload, with the right field names and value types, instead of waiting on a backend team or hand-writing a fake JSON file from scratch.

The same output works well for populating a UI with sample content while you build it. A table, a list of user cards, or a settings form reads very differently when it is filled with varied names, dates and prices instead of the same "asdf asdf" text repeated in every row, and that makes it much easier to spot layout problems, like a long company name that wraps awkwardly or a negative price that was not expected.

What is in the generated data

Names, cities, countries, street names, company names and job titles are drawn from bundled word lists and combined at random, so they read naturally without ever pointing to a real person, business or address. Phone numbers always use the (555) exchange, the block NANP reserves for fiction and testing, so there is no chance of generating a number that could reach a real subscriber. Emails are built the same way, from random name and domain combinations, and every field in a record is randomized independently of the others, so a record's name and email are not derived from each other.

Numeric and date fields respect the ranges you set: an Integer or Float stays within its configured min and max, a Date falls inside the range you pick, and an Enum value is always one of the exact options you typed in. A UUID is a standard version-4 identifier, and the Sentence / Paragraph type produces short blocks of classic lorem-ipsum-style filler text for fields where the exact wording does not matter.

Frequently asked questions

Related: JSON Formatter · UUID Generator · Lorem Ipsum generator.