JSON to SQL Converter

JSON in, SQL out: one store order

Here is a two-row order dump the Sample button loads. Left side is the source. Right side is what Bulk INSERT mode produces for a table named orders.

[{"id": 1042,"sku": "MUG-TEAL","qty": 2,"paid": true,"note": null},{"id": 1043,"sku": "TOTE-NAVY","qty": 1,"paid": false,"note": "gift wrap"}]

INSERT INTO orders (id, sku, qty, paid, note) VALUES
(1042, 'MUG-TEAL', 2, 1, NULL),(1043, 'TOTE-NAVY', 1, 0, 'gift wrap');

Booleans become 1 / 0. Null stays NULL. Strings get quote-escaped. Integers stay bare numbers.

What happens under the hood

Conversion runs entirely in your browser tab. Nothing uploads.

1. Repair, then parse. A repair pass fixes common JSON issues (trailing commas, missing quotes) before JSON.parse runs. Broken input shows the error under the left panel instead of silent failure.
2. Collect columns. For an array, every key across every object becomes a column. A missing key on one row inserts as NULL.
3. Infer types for CREATE TABLE. Numbers become INT or DECIMAL, booleans become BOOLEAN, short strings become VARCHAR(255), longer ones become TEXT. Nested objects and arrays stringify into TEXT.
4. Emit SQL for the mode you picked. One INSERT per row, a CREATE TABLE plus inserts, or a single multi-row INSERT.

Need the JSON cleaned before conversion? Run JSON Fixer or JSON Beautifier first, then paste the result here.

Where this converter stops

No foreign keys, no indexes, no primary key hints. The CREATE TABLE output is a starting scaffold, not a production schema.

• Nested objects and arrays land as escaped JSON text inside one column. There is no automatic JOIN or child-table split.
• Dialect quirks (Postgres TRUE/FALSE, SQL Server bracket identifiers, SQLite affinity) stay out of scope. Output targets common MySQL-style backticks and 1/0 booleans.
• Multi-megabyte payloads slow the Ace editors long before the converter itself fails. Split large dumps.
• Date-looking strings stay strings. No automatic DATE or TIMESTAMP casting.

Prefer a spreadsheet intermediate? JSON to CSV keeps columns flat without inventing SQL types. Prefer YAML configs over table rows? Use JSON to YAML instead.