Free CSV to JSON Converter

What Is CSV to JSON Conversion?

CSV (Comma-Separated Values) is a flat, two-dimensional format — rows and columns, nothing more. JSON (JavaScript Object Notation) is a hierarchical format that can represent nested structures, arrays, and rich types like numbers and booleans.

Converting CSV to JSON is a standard step when moving data between worlds: from a spreadsheet into a web application, from a database export into an API, from an analytics tool into a modern frontend.

The basic shape

Each CSV row becomes a JSON object. Each column header becomes an object key:

name,age,city
Alice,30,New York
Bob,25,London

Becomes:

[
  { "name": "Alice", "age": 30, "city": "New York" },
  { "name": "Bob",   "age": 25, "city": "London" }
]

Note that 30 and 25 are JSON numbers, not strings — the type inference detected them automatically.

RFC 4180 — the CSV standard

CSV looks trivial until you hit real data. The RFC 4180 standard defines the rules that make it parseable:

The parser in this tool implements the full standard. It does not use regex or naive `split(',')` — it walks the input character by character, tracking whether it's inside a quoted field, handling escapes, and correctly distinguishing a delimiter inside a quote from a delimiter outside one.

Delimiter auto-detection

CSV files come in several flavours:

The auto-detector tries each candidate on the first few lines, counts how consistently it produces the same number of columns, and picks the best. If it's wrong, you can override manually.

Type inference — number, boolean, null

CSV stores every value as a string. But "30" and 30 are meaningfully different in JSON — the first is a text value, the second is a number you can do arithmetic on. Type inference converts cells that clearly look like JSON types:

The rules are conservative: only unambiguous matches are converted. Anything ambiguous — leading zeros, thousands separators, trailing dots — stays as a string. You can turn off type inference entirely if you need every cell as a string.

Array of objects vs array of arrays

Two output shapes are common, and different consumers want different ones:

// Array of objects — the standard for APIs and databases
[
  { "name": "Alice", "age": 30 },
  { "name": "Bob",   "age": 25 }
]

// Array of arrays — preserves the raw tabular shape
[
  ["name", "age"],
  ["Alice", 30],
  ["Bob", 25]
]

Array of objects is what nearly every API and database expects. Array of arrays is useful when column names are unknown, when you're feeding data into a spreadsheet engine, or when you want the smallest possible JSON output.

Common use cases

Handling edge cases

• Duplicate headers. A second column called name becomes name_2, a third becomes name_3. Empty headers become col_1, col_2 by position.
• Ragged rows. Rows with fewer cells than headers get empty strings for the missing values. Rows with more get generic col_N keys in object mode.
• Trailing newline. Ignored — no spurious empty record at the end of the file.
• BOM. A leading UTF-8 byte-order mark is stripped automatically.
• No header row. Turn off "First row is header" to generate generic col_1, col_2 keys.

Privacy and security

This CSV to JSON converter runs entirely in your browser. Parsing, type inference, JSON serialisation and download all happen locally. Your CSV never leaves your device — safe for customer records, financial data, internal exports and any other sensitive content.

Frequently Asked Questions

Related Developer Tools