Skip to content

← All tools

JSONL Converter & Validator

Convert between JSONL (newline-delimited JSON) and a JSON array, and validate every record with the exact line number of any error — handy for preparing LLM fine-tuning datasets. Everything runs in your browser; nothing you paste is uploaded.

What is JSONL? Read the guide →

0 non-empty lines

How to use the JSONL converter

  1. Paste a JSON array to turn it into JSONL (one object per line), or paste JSONL to turn it back into a formatted JSON array.
  2. The tool detects the direction and converts as you go.
  3. Copy the result. Everything runs in your browser, so large exports stay private.

JSON vs. JSONL (and NDJSON)

A JSON file is a single document, usually one big array: [{"id":1},{"id":2}]. JSONL (JSON Lines), also called NDJSON, puts one complete JSON object on each line with no enclosing array and no commas between records. The two hold the same data; they differ in how they are read.

When to use each

Things to watch

In JSONL each line must be valid JSON on its own, so an object cannot span multiple lines. Blank lines are ignored. When converting an array to JSONL, every top-level element becomes one line, so the input needs to be an array of records rather than a single object.

Where you'd use this

Preparing and checking fine-tuning datasets, and working with the line-delimited exports that logging and analytics systems produce.

For example: A fine-tuning upload is rejected with "invalid format on line 1,204". The validator points at the exact line: a raw newline inside an assistant message that was never escaped, in a file of 5,000 otherwise valid examples.

Frequently asked questions

What is JSONL?

JSONL (JSON Lines) stores one JSON object per line. It is the standard format for streaming datasets, LLM fine-tuning files and log pipelines because each record can be read independently.

Can it convert between a JSON array and JSONL?

Yes. It converts a JSON array to line-delimited JSONL and back, and validates that every line is well-formed — handy before uploading a training set.

Is my data uploaded?

No. Conversion and validation happen entirely in your browser.

Related tools