CSV to JSON

• First row is the header: header names become the keys. Empty names become column_1, column_2, and duplicates get _2, _3 so no value is lost.
• Detect numbers and true/false: 42 and 3.14 become numbers, true and false become booleans and null becomes null. Values with a leading zero (ZIP codes, phone numbers) and numbers over 15 digits stay strings, because converting them would change the data.
• Empty cells can become empty strings or null, whichever your code expects.
• Nest "a.b" headers: a column called address.city becomes {"address": {"city": ...}}, which reverses what JSON to CSV does.
• Array of arrays keeps the table as rows of values, useful for charts and compact payloads.

The delimiter is detected automatically (comma, semicolon, tab or pipe), and quoted fields with commas or line breaks inside them are parsed correctly. Files are decoded with automatic encoding detection, including UTF-8 with or without BOM, UTF-16, Korean CP949 from Excel and Japanese Shift_JIS, and you can override it. Big files are converted in a background thread, so the page keeps responding while a file with hundreds of thousands of rows is processed. The text box shows the beginning of very large results, while Copy and Download always include everything.

For a spreadsheet instead of JSON, use CSV to Excel. To validate or reformat existing JSON, use the JSON Formatter.