Online test data generator · live · no signup

A free online test data generator you can try in your browser.

Edit a real database schema, generate foreign-key-consistent rows, and export SQL, JSON or CSV, all in the sandbox with no account and nothing to install. When a child row references a parent, that parent exists.

Open the live sandbox Create a free account

Runs in your browser · EU-hosted · zero trackers

From schema to exported rows in three steps

The sandbox opens with a small webshop schema, categories, products, customers, orders and order_items, already wired up with foreign keys. Change anything and regenerate.

  1. Edit the schema. Rename tables, switch column types, add or drop columns, draw a foreign key by dragging between two columns.
  2. Generate data. Pick how many rows per table and generate. Order items point at orders and products that actually exist, so the dataset is referentially complete.
  3. Export it. Download SQL inserts, JSON or CSV and load it into Postgres, MySQL, SQLite or wherever your tests run.

What makes the generated data realistic

FK-consistent

Every reference resolves

The generator orders tables by their dependencies and fills foreign keys from real parent IDs. No dangling references, no constraint violations on import.

column-aware

Values match the column

email gets emails, first_name gets names, price gets sensible amounts, created_at gets timestamps. The column name drives the value.

per-column hints

Override any column

Set a semantic type (first name, city, IBAN, status and more), a list of allowed enum values, or a numeric min and max range, so a column fits your domain instead of a generic guess.

export

SQL, JSON or CSV

Take the data anywhere. SQL inserts for a quick seed, JSON for a mock API, CSV for a spreadsheet or a bulk import.

Honestly: for three hard-coded rows behind a button, a static JSON file is simpler. The sandbox earns its place when your schema is relational, several tables linked by foreign keys, and you want data that imports cleanly and behaves like a real database.

Not flat fake rows, a database that holds together

Most online generators hand you one table of unrelated rows. The moment your app joins two tables, that data falls apart. SeedBase was built schema-first: it was tested against a real 20-app Django project with 226 tables, where a generator that ignores relationships is useless.

SeedBase vs Mockaroo SeedBase vs Faker

FAQ

Is the test data generator free to use without signing up?

Yes. The sandbox runs in your browser with no account and nothing to install. A free account only adds saved projects, larger volumes and pushing straight into a database.

What can I do in the sandbox?

Edit the schema (tables, column types, relations), set per-column hints (semantic type, enum values, min/max range), generate sample rows and export them. Every foreign key resolves to a real parent row.

How is this different from Mockaroo or Faker?

Mockaroo generates flat rows and Faker single values; you wire up the relationships yourself. SeedBase reads the whole schema and generates a referentially consistent database. See the comparison →

Does the data look realistic?

Column names drive the values, and you can override any column with an explicit semantic type, an enum list or a numeric range. More on the generator →

Can I use the generated data in my own database?

Yes. Export as SQL, JSON or CSV and load it into Postgres, MySQL, SQLite or anything else. With a free account you can push directly into a connected database. See the SQL, Django and Prisma guides.

Open the sandbox and generate a dataset now

No account, no credit card. Edit the schema, generate foreign-key-consistent data and export it in under a minute. Create a free account when you want to save projects and scale up.

  • FK-consistent
  • No signup to try
  • Export SQL, JSON, CSV
  • EU-hosted
Open the live sandbox

Related: test data from SQL · Django · Prisma  ·  mock REST API