SPARKPOND / JSON TO CSV

Convert JSON to CSV Online

Inspect the flattened schema before committing to column names. Nested arrays use an explicit selected policy instead of an irreversible automatic guess.

✓ Processed locally✓ No account✓ Original unchanged

LOCAL WORKSPACE

Open JSON or NDJSON

JSON, NDJSON. of zet ze hier neer.

Download the example file ↓

Focused purpose

What this tool does

Convert JSON arrays or NDJSON to CSV locally. Flatten nested objects with path notation, review arrays and missing keys, then download without an account.

  • ✓ Object arrays and NDJSON
  • ✓ Nested object flattening
  • ✓ Path-based headers
  • ✓ Missing-key alignment

Workflow

How to use it

  1. 01

    Open a JSON array or NDJSON stream.

  2. 02

    Review flattened paths and array warnings.

  3. 03

    Choose delimiter and export the table.

Available controls

Available options

  • Choose dot or custom path notation.
  • Serialize, join or reject nested arrays.
  • Select comma, semicolon or tab output.
  • Review the union of all keys.

Nested people objects

Example before and after

Nested properties and a missing city need a stable flat schema.

Before

[{"id":"0001","person":{"name":"Ada","city":"Paris"}},{"id":"0002","person":{"name":"Linus"}}]

After

id,person.name,person.city
0001,Ada,Paris
0002,Linus,

Expected result: Dot paths become columns and the missing key becomes an empty field.

Download example ↓

Practical scenarios

Common use cases

Open API data in a spreadsheet.

Flatten event logs.

Turn NDJSON exports into tabular files.

HONEST LIMITS

Important limitations

  • iDeep or heterogeneous arrays do not have one universal tabular representation.
  • iObject key order does not define the final schema; the preview does.

FAQ

Questions about this tool

Can nested JSON be converted?

Yes. Nested objects flatten to path-based columns shown before export.

What happens to nested arrays?

Choose to serialize them as JSON, join simple values or stop and review.

Can I paste NDJSON?

Yes. Each non-empty line is parsed as one object and aligned to the union schema.