Mock Data Generator for JSON, CSV & SQL
Generate fake names, emails, addresses, and other test data from a custom schema, exported as JSON, CSV, or SQL inserts.
Define a schema with named columns and 21 field types (names, emails, addresses, UUIDs, dates, numbers, and more), then generate up to 1000 rows of realistic fake data with the Faker library. Switch the output between a pretty-printed JSON array, a quoted CSV file, or ready-to-run SQL INSERT statements, and copy or download the result.
Seeding a database, demoing an API, or filling a UI prototype all need plausible records, and writing them by hand wastes time. This tool builds them from a schema you define: each column gets a name and one of 21 field types, covering identity fields (full, first, and last names, emails, phone numbers, job titles), location fields (street addresses, cities, countries), and technical fields (UUIDs, integers with a custom min and max range, decimals, booleans, past and future dates, sentences, paragraphs, URLs, IP addresses, hex colors). The default schema (id, name, email, city) is ready to generate immediately, and you can add or remove columns to match your real table.
Values come from the Faker library, the same package used in countless test suites, so names look like names and addresses look like addresses instead of random character strings. The library (about 3 MB) is downloaded on your first generate and cached, so later runs are instant. Each click of Regenerate produces a fresh random set with the same schema.
Output comes in three formats. JSON gives a pretty-printed array of objects with real number and boolean types. CSV quotes and escapes every field so commas and quotes inside values cannot break the file. SQL produces one INSERT statement per row against a table name you choose, with single quotes escaped and identifiers quoted when needed. Copy the result or download it with the matching file extension.
- 1
Define the schema
Name each column and pick a field type from the dropdown. Integer columns expose min and max inputs. Add or remove columns until the shape matches your target table or API.
- 2
Set count and format
Choose how many rows to generate (1 to 1000) and pick JSON, CSV, or SQL output. The SQL tab adds a table name input for the INSERT statements.
- 3
Generate and export
Click Generate to produce the data, Regenerate for a fresh random set, then copy the output or download it as a .json, .csv, or .sql file.
Seed a development database
Generate 500 rows of SQL INSERT statements matching your users table and run them against a local Postgres or MySQL instance.
Stub an API response
Produce a JSON array of fake customer objects to serve from a mock endpoint while the real backend is still being built.
Test a CSV import pipeline
Create a CSV with quoted fields, commas inside addresses, and edge-case values to verify your importer parses real-world files correctly.
Fill a UI prototype
Generate realistic names, job titles, and companies so a table or card layout can be reviewed with believable content instead of lorem placeholders.
Is the generated data sent anywhere?
No. The Faker library runs inside your browser tab, and the schema, settings, and generated rows never leave your machine. Nothing is uploaded or logged.
Is the data based on real people?
No. Faker assembles values from word lists and patterns, so any resemblance to a real person, address, or company is coincidental. It is safe to use in demos and screenshots.
What SQL dialect do the INSERT statements use?
Standard SQL: single-quoted strings with quotes doubled for escaping, TRUE and FALSE for booleans, and unquoted numbers. The statements run on PostgreSQL, MySQL, SQLite, and most other databases without changes.
How are dates formatted?
Past date and Future date fields produce ISO 8601 timestamps such as 2025-11-03T14:22:08.510Z. Past dates fall within the last year and future dates within the next year.
Why is there a 1000 row limit?
Generation happens in your browser, and rendering very large outputs makes the page sluggish. For bigger datasets, generate in batches of 1000 and concatenate the files, or download several SQL batches and run them in sequence.