How to Convert CSV to JSON

CSV looks like a simple grid until a value contains a comma. Then the same file that opened correctly in a spreadsheet becomes a broken row in a naive converter. The difference between a useful conversion and a corrupted dataset is usually quote handling and a clear decision about types.

The basic mapping

A CSV file with a header row:

name,role,active
Ada,admin,true
Sam,editor,false

becomes:

[
  { "name": "Ada", "role": "admin", "active": "true" },
  { "name": "Sam", "role": "editor", "active": "false" }
]

The first row becomes the object keys. Each later row becomes one object. That is the whole model.

Notice that active is a string. CSV has no type system, so a converter should not guess that true is a boolean or that 00123 is a number. Keeping values as strings is usually safer, especially for ZIP codes, phone numbers, and IDs with leading zeros.

A reliable conversion workflow

1. Export a clean CSV. Remove title rows, merged cells, and formulas. A header row should be the first row.
2. Choose the delimiter. Comma is common; semicolon and tab appear in locale-specific exports.
3. Check quoted values. A field such as "New York, NY" must stay one value, not become two.
4. Keep IDs as strings. Leading zeros disappear if a tool coerces the value to a number.
5. Check the row count. Compare the array length with the number of data rows in the spreadsheet.
6. Validate the JSON. A converter should produce valid JSON; if it does not, the input probably has an unclosed quote.

CSV to JSON parses comma, semicolon, and tab delimiters in the browser. The file is not uploaded.

What CSV cannot represent well

• Nested structures. A cell containing JSON can be kept as a string, but it is not expanded automatically.
• Multiple header rows. Flatten them into one header before converting.
• Formulas. Export values, not formulas.
• Dates and time zones. A date may be a string, a serial number, or a locale-specific format. Normalize it before conversion.
• Very large files. A browser tab has memory limits; split large exports into batches.

When to use a different tool

If you need to edit the data, use a spreadsheet. If you need to query it, load it into a database. If you need a JSON schema, validate the converted output rather than assuming the CSV was clean.

For JSON that already exists, JSON Formatter is the better first step. For a page’s metadata, Meta Tag Generator handles a different kind of structured output.

The short version

Convert rows to objects, keep values as strings, and check quoted commas and row counts. A fast conversion is not useful if it silently changes an ID or splits an address.