NDJSON / JSON Lines Formatter & Validator
Validate and pretty-print newline-delimited JSON, or convert to/from a regular JSON array
Output will appear here…]. Blank lines are skipped.About this tool
The ToolNinja NDJSON / JSON Lines Formatter validates and pretty-prints newline-delimited JSON — one independent JSON value per line, rather than one big array — the format behind streaming API responses, log files, and increasingly, token-by-token LLM output. Each line is parsed and validated independently, so a single malformed line is reported with its exact line number rather than failing the whole input the way a single JSON.parse() over the full text would. A Convert mode turns valid NDJSON into a regular JSON array (and back) for the cases where you need the data in one format or the other — exporting NDJSON logs into a JSON array for a script that expects one, or splitting a JSON array into NDJSON for a streaming pipeline. Everything runs 100% in your browser — your log data or API response never leaves your machine.
When to use it
- →Validating a log file or API response that uses NDJSON/JSON Lines format, with exact line numbers for any malformed entries
- →Converting NDJSON log output into a regular JSON array for a script or tool that doesn't support the line-delimited format
- →Converting a JSON array into NDJSON for a streaming pipeline or bulk-import API that expects one record per line
- →Pretty-printing minified NDJSON log lines to read them more easily during debugging
Tips
- ◆Blank lines are silently skipped — they're a common artifact of how NDJSON files get generated and aren't treated as errors.
- ◆Converting to a JSON array requires every line to be valid first — fix the lines listed as errors before the array conversion will produce output.
- ◆NDJSON and JSON Lines (.jsonl) are the same format under different names — both mean exactly one JSON value per line, no trailing comma, no wrapping array.
Frequently asked questions
What's actually different between NDJSON and a regular JSON array?
A JSON array must be fully received and parsed as one unit — you can't know the array is 'done' until you see the closing ]. NDJSON has no wrapping structure at all; each line is independently complete and parseable the moment it arrives, which is exactly why streaming APIs and log pipelines prefer it — a consumer can process line 1 while line 50,000 is still being generated.
Why do LLM streaming APIs often use a line-delimited format?
Token-by-token generation produces output incrementally, and a consumer needs to process each chunk as it arrives rather than waiting for the full response — the same requirement NDJSON was designed for. Many LLM streaming endpoints use NDJSON directly, or a closely related line-delimited event format (like text/event-stream), for this reason.
Can NDJSON have trailing commas or wrap everything in brackets?
No — each line must be a complete, standalone JSON value with no trailing comma and no enclosing array brackets. A file that starts with [ and separates entries with commas is a regular (possibly pretty-printed) JSON array, not NDJSON, even if it has one value per line.