Paste a JSON document or import a .json file.
Define row selection, nested-array handling, and the CSV serialization policy.
Validation results and export actions appear here.
Add a JSON document, validate it, then review the row mapping and CSV output.
Convert JSON to CSV after choosing which array becomes the rows. Review flattened paths, nested-array handling, blank values, long IDs, and spreadsheet-sensitive cells before downloading the CSV locally.
A useful flat file requires three explicit choices: which JSON array defines the rows, how nested arrays should be represented, and how blank or spreadsheet-sensitive values should be written. The preview exposes the tradeoffs before export.
Paste JSON or choose a local file. The parser identifies candidate record arrays, blocks duplicate object keys, and retains long numeric tokens before they can be rounded.
Choose the array that becomes rows. Keep nested arrays as JSON, join scalar values, or expand one array; unrelated arrays are never multiplied into an accidental cross product.
Inspect flattened columns, missing fields, nulls, empty strings, long IDs, and formula-like values. Convert JSON to CSV only when the preview reflects the intended table.
{
"items": [
{ "id": 1234567890123456789, "tags": ["new", "priority"], "note": null },
{ "id": 42, "tags": [], "note": "" }
]
}Choose items as the record path, then keep tags as JSON text or join the scalar values. The preview distinguishes null from an empty string and preserves the 19-digit ID; the downloaded CSV necessarily writes both blank states as empty fields.
Return to the converterDirect answers about nested fields, arrays, blank values, long IDs, and spreadsheet protection.
Paste or select the file, validate it, choose the records array and an array rule, then inspect the flattened preview. Download only after the row count and path-based columns match the source records you intended to export.
Nested objects become path-based columns, while literal keys containing dots are escaped so they remain distinct. This prevents two different source fields from being merged into one output column.
Arrays can remain as JSON text, join when every value is scalar, or expand one selected array into rows. Unrelated arrays are never multiplied automatically because that would invent combinations absent from the source.
The preview distinguishes null, a missing field, and an empty string so you can review their different source states. The downloaded file writes all three as blank fields because CSV has no native null or missing-value type.
Long integer tokens retain every source digit, and formula-like strings can be prefixed for safer spreadsheet use. The optional report records how many cells were protected, although spreadsheet software may still infer other value types.