NDJSON and CSV: what actually changes
NDJSON is one JSON object per line, so a huge file can be streamed rather than parsed whole. CSV is comma-separated plain text — one table, no types, readable by everything.
Records are read out of the NDJSON with their structure intact and written as CSV, so this is a real parse rather than a text substitution — quoting, escaping and character encoding are all handled for you. The CSV you get back keeps the header row and every value as text, and gives up formulas, formatting, multiple sheets and column types. The thing to check before you rely on the output is typing: dates, leading zeros and very large integers are where formats disagree.
How it goes3 steps
-
Add your NDJSON files
Drop them onto the box above or click Select File. Add as many .ndjson files as you like — the whole batch converts in one go, and duplicates are ignored.
-
Confirm CSV as the output
CSV is already selected as the target format on this page. Check the delimiter and encoding options if your data is unusual.
-
Download your CSV
Hit Convert and grab the finished CSV files individually or as a single ZIP. Everything is deleted from the server once you have them.
Format facts
What each side of this conversion is good at, so you know what you are trading away.
Source
.ndjson
NDJSON
One JSON object per line, so a huge file can be streamed rather than parsed whole.
- Compression
- Lossless
- Transparency
- Not supported
- Keeps
- Record structure while staying line-addressable
- Gives up
- The enclosing array — each line stands alone
Result
.csv
CSV
Comma-separated plain text — one table, no types, readable by everything.
- Compression
- Lossy
- Transparency
- Not supported
- Keeps
- The header row and every value as text
- Gives up
- Formulas, formatting, multiple sheets and column types
Send NDJSON somewhere else
Every other format a NDJSON file can leave as.
From .ndjson
6 targets