CSV to JSON Converter

How to convert CSV to JSON

1. Paste the CSV, drop a .csv or .tsv file onto the input, or press Upload. Sample loads an example.
2. Leave First row is the header ticked if the first line holds column names. The JSON updates as you type.
3. Choose Array of objects (one object per row) or Array of arrays, and the JSON indentation.
4. Press Copy or Download .json. To check or tidy the result further, format and validate the JSON, or browse it in the JSON viewer.

CSV to JSON example

A small product list, with a quoted name that contains a comma:

sku,name,price,in_stock,discontinued
A-100,"Desk, oak",129.5,true,null
A-101,Lamp,19,false,
A-102,Chair,007,true,false

becomes an array of objects, one per row, keyed by the header:

[
  {
    "sku": "A-100",
    "name": "Desk, oak",
    "price": 129.5,
    "in_stock": true,
    "discontinued": null
  },
  {
    "sku": "A-101",
    "name": "Lamp",
    "price": 19,
    "in_stock": false,
    "discontinued": ""
  },
  {
    "sku": "A-102",
    "name": "Chair",
    "price": "007",
    "in_stock": true,
    "discontinued": false
  }
]

Notice what happened to the values: 129.5 and 19 became numbers, true/false became booleans, null became JSON null, the empty cell stayed "", and 007 stayed a string so the leading zeros survive.

Numbers, true/false and null

CSV has no types: every cell is text. With Detect numbers, true/false, null ticked, the converter types only the values that are safe to convert:

• Plain numbers like 42, -3.5 and 1e6 become JSON numbers.
• true/false (also TRUE, False) become booleans, and null becomes null.
• Leading zeros, +1, 1,000, dates and integers above 9,007,199,254,740,991 stay strings.

Untick it to keep every value as a string, which is what some importers expect:

[
  {
    "sku": "A-100",
    "price": "129.5",
    "in_stock": "true"
  }
]

Commas, semicolons and tabs

The delimiter is detected automatically (comma, semicolon, tab or pipe). Many European spreadsheets save semicolon-separated files with a decimal comma, which stays text rather than being misread:

[
  {
    "name": "Desk, oak",
    "price": "129,50"
  },
  {
    "name": "Lamp",
    "price": 19
  }
]

Quotes, line breaks and messy CSV

The parser follows RFC 4180: a field in double quotes can contain commas, line breaks and doubled quotes (""). This CSV

id,comment
1,"She said ""ship it""
then left"

converts to

[
  {
    "id": 1,
    "comment": "She said \"ship it\"\nthen left"
  }
]

It also copes with the usual problems in exported files:

• A byte order mark at the start (common in Excel exports) is removed, so the first key isn't "id".
• Blank lines are skipped.
• Uneven rows are padded or given extra keys, with a warning.
• A quote that is never closed stops the conversion and shows where it starts. For example: A quoted field is never closed: a " opens a field but no matching " ends it. (line 2, column 3). Press Show in input to jump to it.

Array of objects or array of arrays?

Array of objects is easiest to work with in code, because you read values by name (row.price). Array of arrays is smaller and keeps the header as the first row, which suits charting libraries and bulk inserts:

[["sku","price"],["A-100",129.5],["A-101",19]]

Convert CSV to JSON in Python or JavaScript

• Python: json.dump(list(csv.DictReader(open("in.csv", newline="", encoding="utf-8-sig"))), open("out.json", "w"), indent=2). Every value stays a string, so convert types yourself.
• JavaScript: Papa.parse(text, { header: true, dynamicTyping: true, skipEmptyLines: true }).data with the Papa Parse library, which this tool also uses for parsing. Note that dynamicTyping turns 007 into 7.

Need the reverse? The JSON to CSV converter turns JSON back into a spreadsheet-ready CSV.

Frequently asked questions