CSV to JSON Converter

Converting CSV to JSON correctly

CSV looks simple enough to split on commas and line breaks, but real files break that approach quickly: a name like "Hopper, Grace" contains the delimiter, an address field contains a line break, and a quote inside a quoted field is written as two quotes (""). This converter uses a proper RFC 4180 parser that handles all of these, plus CRLF and LF line endings, a trailing newline and the byte order mark Excel adds to UTF-8 exports.

The delimiter is detected automatically by checking which of comma, semicolon, tab and pipe splits the first rows into a consistent number of columns — useful for European Excel exports, which use semicolons. With the first row as the header, each line becomes an object whose keys are the column names; duplicate or empty header names are renamed and reported rather than silently overwriting data.

Every CSV value is text. Type detection turns 42, 3.14, true and null into real JSON numbers, booleans and null, while keeping values like the zip code 02134 and 19-digit IDs as strings so nothing is corrupted. Rows with too few or too many fields are flagged with their line numbers.

How to use the CSV to JSON Converter

1. Paste CSV, open a .csv or .tsv file, or click Sample.
2. Leave the delimiter on Auto-detect or choose one, and untick First row is the header if your file has no header.
3. Choose the output shape — array of objects, array of arrays, an object keyed by a column such as id, or an object of column arrays — and how empty cells should appear.
4. Review any warnings about mismatched rows or header names, then copy the JSON or download data.json.

Example

With type detection on, this CSV:

id,name,active,zip
1,"Hopper, Grace",true,02134

// becomes
[
  {
    "id": 1,
    "name": "Hopper, Grace",
    "active": true,
    "zip": "02134"
  }
]

Choosing an output shape

• Array of objects — the usual choice for APIs, fixtures and importing into a database: [{"id": 1, "name": "Ada"}].
• Array of arrays — compact and order-preserving, with the header as the first row; good for charting libraries and data tables.
• Object keyed by a column — {"1": {...}, "2": {...}} for fast lookups by ID. Repeated key values are reported because later rows replace earlier ones.
• Object of column arrays — {"id": [1, 2], "name": ["Ada", "Grace"]}, the columnar layout used by many plotting and dataframe tools.

Common CSV problems and what they mean

• Every row has a different column count: the delimiter is probably wrong; try a different one or check the detected value.
• "A quoted field is never closed": a field opens with " but the closing quote is missing, so the parser read the rest of the file into one cell. Look at the reported line.
• Unexpected quote in an unquoted field: a value like 5" screen should be written as "5"" screen". The quote is kept as-is and reported.