JSON Splitter — Record Count or File Size

About this tool

Split JSON containing record arrays, or NDJSON with one object per line, by record count or file size. JSON automatically selects the largest record array, with manual path selection available. Large integers, decimals and source number notation remain unchanged. Output stays valid JSON / NDJSON, and the entire file is validated before saving. Actual encoded bytes determine the size limit. Processing runs entirely in your browser (no data is sent).

Input file and format

Select a file.

Records to split

Choose a file to see record candidates.

The array with the most records is chosen, preferring a shallower path for ties. Inspect or change the candidate, or stop a lengthy scan and choose the root or a manual path.

Supported examples: [{"id":"001","name":"Alice"},{"id":"002","name":"Bob"}], [["001","Alice"],["002","Bob"]] and NDJSON. For nested data such as {"data":{"items":[...]}}, choose /data/items. Do not mix object and array rows in one target array. NDJSON requires objects. Arrays of standalone numbers or strings are not supported.

The detected input format will appear after file selection.

Input preview (encoding and structure)

      

Only the beginning is shown. Full validation checks the entire file before splitting.

Split limits and output format

Split and save settings

Size limits include BOMs, line endings, brackets and tags. 1 MB = 1,000,000 bytes.

For JSON destinations without specified encoding requirements, UTF-8 without BOM is the usual choice.

Choose a file and review the preview and split settings.

How to use

  1. Select the JSON file and check the preview's encoding and structure. Check the automatically chosen record path and count, or select a candidate/manual path. For NDJSON / JSON Lines, set the matching data format.
  2. Set the maximum file size or record count according to your import limits. Formatting line breaks do not count as rows.
  3. Choose an output encoding and saving method, then select Validate and split.
  4. Folder saving creates a dedicated subfolder, with names such as original-name_part_000001.json (or .ndjson for NDJSON) are numbered sequentially. In download mode, confirm each file before proceeding.

Processing behavior and limits

  • Sizes are calculated from actual uncompressed output files. Multibyte characters count by their actual encoded size. If the document envelope and one record alone exceed the limit, processing stops before saving and reports the required size.
  • Each output array is enclosed in [ ... ] with commas and LF line endings between records. Outer objects and shared metadata are retained for each output. NDJSON uses LF between records. Original strings, large integers, decimals, exponent notation, -0 and key order are preserved without converting types.
  • Missing fields and nested values are preserved. The first element is not treated as a header or duplicated in each output. An empty [] produces one output; an empty file produces none.
  • Automatic detection scans the entire input. Before writing any output, validation covers record boundaries, surrounding document data, syntax, encoding, representability and size, including the file's end. Shared metadata repeated in outputs is limited to 16 MiB.
  • Records are processed without loading the entire file into memory. Each record is limited to 64 MiB, and individual downloads to 128 MiB per output file. For large files, use folder saving in a supported desktop browser.
  • On error or cancellation, only the dedicated folder created for this run is deleted. If cleanup fails, its location is shown. Downloads cannot be deleted automatically; discard incomplete results rather than importing them.