Mockaroo's UI is excellent, but configuring 50 fields by hand for every new table gets old. MetaDataLoop reads your CREATE TABLE statement directly: it infers what each column probably contains from the column name (email → email, full_name → person name, phone → phone) and the type (NUMERIC(10,2) → currency-shaped decimal, DATE → realistic recent date), and produces ready-to-use rows in seconds.
What you get
✓ INCLUDED
Skip the 50-field UI — your CREATE TABLE is already the spec
✓ INCLUDED
Smart inference: emails, names, phones, addresses, IPs, URLs, lorem text
✓ INCLUDED
Dialect-aware output: Postgres, MySQL, Snowflake, SQL Server INSERT
✓ INCLUDED
Bundled with the rest of MetaDataLoop's data-engineering toolkit
✓ INCLUDED
Free, no signup, no row-limit gotcha
Side-by-side comparison
Feature
Mockaroo (free)
MetaDataLoop
Signup required
Yes
No
Configure each field
Manual
Auto from DDL
Read CREATE TABLE input
No
Yes
Free row limit
1,000
1,000
CSV / JSON / SQL output
Yes
Yes
Snowflake INSERT syntax
Generic
Yes (dialect-aware)
Schema diff & migrations
No
Yes
Schema → Mermaid / DBML
No
Yes
Frequently asked questions
Can I customize how a single field is generated?
Currently inference is automatic — column name + type drive the value. For fully custom field-by-field configuration, Mockaroo is still the more flexible tool.
What about referential integrity?
Generate the parent table first, then use those PKs as the FK column values when generating the child table. Cross-table FK auto-wiring is on the roadmap.
Can I generate millions of rows?
Up to ~1,000 rows in-browser. For larger volumes, generate the SQL INSERT, then loop in your client or use a generate_series pattern in Postgres/Snowflake.
Try it now
No signup, no install — runs entirely in your browser.
Your privacy matters.
We use a strictly-necessary session cookie to keep you signed in.
With your consent, we may also set analytics cookies to improve the product.
You can change your choice anytime from the footer.
Privacy policy →