Why are all my numbers strings in the output?

CSV carries no type information, so guessing that 30 is a number would break values like 007 or a zip code with a leading zero. Every cell stays a string and typing is left to your code.

Can it read an Excel .xlsx file?

No. An .xlsx is a zipped workbook, not CSV. Save or export as CSV first, then paste the text here.

My file uses semicolons instead of commas.

Replace the semicolons with commas in a text editor, or export from your spreadsheet app with a comma separator.

What happens to a row with too few values?

Missing cells become empty strings and the row numbers are listed above the output, so ragged rows stay visible instead of being silently normalised.

CSV to JSON — Guide

Convert CSV to a JSON array with quoted fields, real newlines and duplicate-column safeguards.

What it is and when to use it

CSV has more conventions than rules. RFC 4180 defines quoted fields, doubled quotes as the escape and CRLF line breaks, but real exports from spreadsheets use semicolons, a UTF-8 BOM or bare LF depending on locale. A converter that just splits on commas quietly corrupts any cell containing a comma or a newline, which is why this tool parses character by character.

Paste CSV and get a JSON array of objects, one per data row, keyed by the header row. Quoted fields, commas and real line breaks inside a cell, doubled quotes, a leading BOM and CRLF endings are all handled.

How to use it

Paste the CSV text into the input box.
Leave "First row holds the column names" on when the first row is a header; turn it off to get column_1, column_2 keys instead.
Copy the JSON result.

A worked example

Load an example CSV

name,role,joined
Ada,engineer,2021-03-02
Grace,lead,2019-07-11

Copy this into the tool, or open CSV to JSON

Things that trip people up

Every value becomes a string. Numbers, dates and booleans are not guessed, so 30 stays "30" instead of becoming the number 30.
A row with more values than the header is not dropped: the extra values are kept under __extra_1, __extra_2 and the row number is reported above the output.
Duplicate column names are renamed (a, a__2) instead of overwriting each other, and the rename is listed so you can fix the source.
An empty header cell becomes column_N so no column is lost.
An unterminated quote is reported as an error with its line number rather than producing a truncated result.

Questions people ask

Why are all my numbers strings in the output?: CSV carries no type information, so guessing that 30 is a number would break values like 007 or a zip code with a leading zero. Every cell stays a string and typing is left to your code.
Can it read an Excel .xlsx file?: No. An .xlsx is a zipped workbook, not CSV. Save or export as CSV first, then paste the text here.
My file uses semicolons instead of commas.: Replace the semicolons with commas in a text editor, or export from your spreadsheet app with a comma separator.
What happens to a row with too few values?: Missing cells become empty strings and the row numbers are listed above the output, so ragged rows stay visible instead of being silently normalised.

Next steps

Open the CSV to JSON tool

Spotted something wrong on this page? Tell us via Contact Us