Will numbers come out as numbers?

Yes. Types are inferred per column and written properly, rather than quoting everything as a string. You can override the inference on any column.

Can I get NDJSON instead of one big array?

Yes — pick NDJSON and you get one record per line, which is what most bulk-load and log endpoints want.

Does the JSON get a BOM?

No. A byte-order mark is helpful in CSV for Excel and malformed in JSON, so DataTray writes it in the first case and never in the second.

Free converter

CSVJSON

CSV to JSON Converter

Going the other way — a spreadsheet export into something an API or a script can consume. DataTray infers the column types so numbers come out as numbers rather than every field being a quoted string.

Windows 10/11 (x64) · Free · No account

Drop in your files — JSON and NDJSON are written without a byte-order mark, which is what parsers expect.

Why convert

Types, and why they matter here

A CSV has no types — everything in it is text. A naive conversion therefore produces JSON where every value is a quoted string, and whatever consumes it has to re-parse all of them. DataTray infers types from the column and writes numbers, booleans and nulls properly, with the inference overridable per column.

Feeding a spreadsheet export into an API or a script.
Producing NDJSON for a log pipeline or a bulk-load endpoint.
Turning a lookup table into a JSON fixture.
Converting a CSV too large for an online tool to accept.

How it works

Convert CSV to JSON in a few clicks

1
Open the CSV
The delimiter and encoding are detected and shown, and you can correct either before reading.
2
Confirm the types
DataTray infers a type per column. Override any of them — this is what stops every value becoming a string.
3
Pick JSON or NDJSON
One array, or one record per line for streaming consumers.

The details

CSV to JSON, at a glance

Output: JSON array or NDJSON / JSONL, one record per line
Types: Inferred per column, and overridable
Encoding: UTF-8, never with a BOM — a BOM in JSON is malformed
Batch: A folder at a time

Questions