How does the tool decide a column is nullable?

When the input is an array of objects, the tool compares keys across every item. A key missing from at least one item, or holding a null value in any item, is generated without a NOT NULL constraint.

What SQL type does it use for strings?

Short strings become VARCHAR with a length rounded up from the longest value seen (up to 255). Strings longer than 255 characters, or columns with no length information, become TEXT.

How are nested objects and arrays handled?

A field holding a nested object or array is generated as a JSON column (JSON in MySQL, JSONB in PostgreSQL, TEXT in SQLite), since these dialects do not have a single flat equivalent for nested structures.

What happens if a column has mixed types across items, like a string in one row and a number in another?

Mixed-type columns fall back to TEXT so no value is lost, since no single native SQL type covers both cases safely.

Does it add a primary key automatically?

Yes, by default an auto-incrementing id column is added as the primary key. Uncheck "Add id primary key column" to generate only the columns found in the JSON.

Is my JSON sent to a server?

No. Parsing and SQL generation run entirely in your browser with plain JavaScript. Your data never leaves your device.

JSON to SQL Schema

Paste JSON and get a matching CREATE TABLE statement, with column types inferred and a MySQL, PostgreSQL, or SQLite dialect picker.

Input JSON

Table name

Add id primary key column

Dialect

SQL Output

Paste JSON and click Generate

Copied!

Summary

JSON to SQL Schema reads a pasted JSON object or array of objects and generates a matching CREATE TABLE statement. When the input is an array, the tool scans every item to infer each column type and marks a column nullable if it is missing from some entries or holds a null value. Column types, quoting, and the primary key clause adjust to the selected SQL dialect: MySQL, PostgreSQL, or SQLite.

How it works

Paste a JSON object or an array of JSON objects into the input panel, or load a sample.
Set a table name and pick a SQL dialect: MySQL, PostgreSQL, or SQLite.
Click Generate to scan the JSON and infer a type for every column.
Columns missing from some array items, or holding null, are generated as nullable.
Review the generated CREATE TABLE statement in the output panel.
Click Copy or Download to save the SQL.

Use cases

Turn a sample API response into a starting database table.
Scaffold a table for a new project from a JSON fixture.
Check which fields in a JSON export are sometimes missing before writing a migration.
Convert a JSON config or seed file into an equivalent SQL table.
Draft a schema for MySQL, PostgreSQL, or SQLite from the same JSON source.
Document the shape of a JSON payload as a SQL table for a teammate.

Frequently Asked Questions

Last updated: 2026-07-14 · Reviewed by Nham Vu