Contact

Tools  /  JSON  /  CSV → JSON

JSON

CSV → JSON

Parse CSV (quoted fields, any delimiter) into an array of JSON objects -- auto-detect delimiter, optional header row, type coercion, dot-path column unflattening into nested objects, duplicate-header renaming, and a column summary.

More JSON tools

JSON Validator JSON ↔ JS Object Literal Array Deduplicator Array Sorter JSON Minifier Array Shuffler

What this tool does

Paste or load CSV and get back an array of JSON objects, one per row, with column headers becoming keys — using an actual CSV parser rather than splitting each line on a comma, which breaks the moment a field contains one.

A real parser, not a naive split

A quoted field can contain the delimiter itself, a line break, or an escaped quote, and none of that should end the field early. This is parsed properly rather than by splitting text on commas, so a value like "Doe, John" stays one field instead of breaking into two. The delimiter itself — comma, semicolon, tab, or pipe — is auto-detected by sampling the first few lines, or can be set explicitly.

Turning flat columns into nested JSON

A CSV can only have flat column names, but the data they represent is often naturally nested. With unflattening turned on, a header like address.city creates a nested address object with a city field, instead of a flat key that happens to contain a dot — recovering structure that CSV's format has no way to express directly.

Two silent-data-loss bugs this avoids

  • Duplicate headers — if two columns share the same name, a naive converter has the second value silently overwrite the first in the resulting object. Here, the second occurrence is renamed instead, so both values survive in the output.
  • A leading byte-order mark — CSV files exported from Excel commonly start with an invisible BOM character, which a naive parser leaves stuck to the front of the first header name, breaking every reference to that column. It is stripped automatically before parsing begins.

Privacy

Parsing and conversion both happen in your browser. Nothing you paste, drop, or load from a file is sent to a server.

FAQ

Frequently asked questions

Is my data uploaded anywhere?
No. Parsing and conversion both run locally in the browser — nothing is sent to a server.
Does it handle quoted fields with commas or newlines inside them?
Yes — parsing follows real CSV quoting rules rather than splitting text on the delimiter, so a quoted field containing the delimiter itself, a line break, or an escaped quote stays one field instead of breaking apart.
What does unflattening dot-path columns do?
A header written as address.city creates a nested address object holding a city field in the output, instead of a single flat key that happens to contain a dot. It recovers the nested structure that CSV itself has no way to represent directly.
What happens if two columns have the same header name?
The second occurrence is automatically renamed rather than silently overwriting the first column's values in the resulting object, which is what a naive converter using the header directly as a key would do.
Why does a CSV exported from Excel sometimes break other converters?
Excel often writes an invisible byte-order-mark character at the very start of the file. A parser that does not account for it ends up with that character stuck to the front of the first header name, breaking every reference to that column. It is stripped automatically here before parsing begins.
What output formats does it support besides plain JSON?
Pretty-printed JSON, compact JSON, newline-delimited JSON (one object per line), or an array of arrays that keeps the header row as the first entry.
Theme Accent