What the JSON looks like
Each row becomes an object whose keys are the column names:
[
{ "order_id": 1001, "customer": "Ava Chen", "unit_price": 89.99 },
{ "order_id": 1002, "customer": "Liam Brooks", "unit_price": 14.5 }
]
Numbers stay numbers: QueryLocal detects each column's type from the data, so 89.99 is exported as a number, not as "89.99". Empty numeric cells become null.
Convert only what you need
The JSON export contains the result of the current query, not necessarily the whole file. Choose and rename fields as you go:
SELECT order_id AS id,
customer AS name,
quantity * unit_price AS total
FROM data
WHERE country = 'USA';
The download always contains every row of the result; the screen only shows the first 1,000.
Column names
Headers are turned into safe keys: accents removed, spaces and symbols replaced by underscores (Date de création becomes Date_de_creation). Use AS in the query to choose any other name.
Other free CSV tools
Frequently asked questions
Can I get JSON Lines (one object per line) instead?
The export is a standard JSON array, which most tools accept. If you need JSON Lines, most languages convert an array in one line of code.
What about nested JSON?
A CSV is flat, so the output is an array of flat objects. Use column aliases with dots (AS "address.city") if the tool you feed expects them.
Does it handle semicolon CSV files from Excel?
Yes, the separator is detected automatically, and decimal commas (12,5) are read as numbers.