CSV to JSON Converter

What does this converter do?

It turns CSV into JSON in your browser, and it makes one decision differently from the usual approach: the type of a column is decided for the column as a whole, never cell by cell.

• Per-column typing. One ZIP code with a leading zero keeps the entire column as text, so you never get 2139 in one row and "02139" in the next.
• It shows what it decided. A panel lists every column with the type chosen, how many values were empty, and the reason a column was held as text.
• Override any column. Force text, number or boolean when you know better than the detection does.
• Delimiter detection that names itself. Comma, semicolon, tab or pipe, with the one it found written on screen and a way to change it.
• RFC 4180 parsing. Quoted fields can contain the delimiter, line breaks and escaped quotes.
• Nested output from dot-notation headers, and nothing ever leaves your browser.

Why per-column typing matters

CSV has no types. Every value in the file is text, so a converter has to decide what each one means. Deciding cell by cell is where data quietly breaks:

Look at the middle column. The same field comes back as a number in one row and a string in another, because each cell was judged alone. That mixed field is worse than losing the zero: it passes a quick glance and breaks the thing that reads it later.

Deciding for the whole column means one value with a leading zero protects every other value in that field.

The delimiter problem nobody warns you about

Excel in most of Europe saves CSV using semicolons. The reason is arithmetic, not preference: in those locales the comma is the decimal separator, so 3,14 is a number and a comma delimiter would make the file impossible to read.

The file still ends in .csv, so a converter expecting commas collapses every row into one long field and produces JSON with a single garbled key. This page detects the delimiter, tells you which one it found, and lets you correct it.

What this tool does not do

• It does not convert dates. 03/04/2026 is March in the United States and April in most of Europe, and nothing inside a CSV says which. A wrong guess corrupts data silently, so dates stay exactly as written.
• It does not read .xlsx files. That is a zipped XML package, not text, and unpacking it is a different problem. Save as CSV first.
• It does not repair broken CSV. When a row has more or fewer fields than the header, it tells you which row, rather than guessing what you meant.
• It does not stream very large files. Everything runs on the page's main thread, so a file of tens of megabytes can make the tab stall. That is a real limit, not a hidden one.
• It does not upload anything. Which also means there is no size cap imposed by a server, only your device's memory.

Is my data private?

Yes. The parser is plain JavaScript running in your own browser, and the file never leaves your device. Nothing is uploaded, logged or stored, there is no account and no email field, and closing the tab clears everything. The JSON appears the instant you paste, because there is nothing to submit and no one to wait for.