Synthetic data is generated instead of collected from real people. For development and QA that means you can fill databases, test forms and demo products without touching personal data, and share datasets with colleagues and vendors freely.
Production copies, anonymised data or synthetic data
| Production copy | Anonymised copy | Synthetic data | |
|---|---|---|---|
| Contains real people | Yes | Indirectly; can be re-identified | No |
| Risk if it leaks | High | Medium | Negligible |
| Needs access to production | Yes | Yes | No |
| Size and shape on demand | No | No | Yes |
| Edge cases on purpose | Only if they happened | Only if they happened | Yes |
Synthetic test data with RNDGen
RNDGen generates synthetic data from a schema you control, entirely in your browser. It is not derived from your production data, so there is nothing to leak, and you decide the size, the shape and the edge cases.
- 240+ field types for people, addresses, companies, finance, internet, products, travel and more.
- Realistic, consistent rows and related tables with real foreign keys, the essence of simulated data.
- Edge cases for QA: Unicode and emoji strings, SQL injection and XSS payloads, empty values.
- Payment test values that cannot charge anyone: Luhn-valid test card numbers, Stripe test cards and valid-format IBANs.
- Every format: CSV, Excel, JSON, SQL, XML and more, or a live mock API.
Learning distributions from a sample of your own data is on our roadmap; follow What's new for updates.
FAQ
Is synthetic data covered by GDPR?
Data generated from a schema, like RNDGen’s, describes no real person, so it is not personal data. Data derived from real records (masked or anonymised copies) can still be personal data if people can be re-identified.
Is synthetic data the same as anonymised data?
No. Anonymised data starts from real records and removes or scrambles identifiers. Synthetic data is created from scratch, so there is nothing to re-identify.
Can RNDGen learn from my real data?
Not today. RNDGen generates data from a schema you define or that AI designs from a description; it never sees your production data. Learning distributions from a sample you provide is on our roadmap.
How much synthetic data can I generate?
Up to 100,000 rows per file without an account, 1,000,000 with a free account and 5,000,000 on Pro and Team, as many files as you need.