JSONL Viewer

What is JSON Lines?

JSON Lines (also called JSONL or NDJSON) is a text format where each line is a complete, independent JSON object. It is the default for streaming records, log files, machine-learning datasets and LLM fine-tuning data, because you can append and process one line at a time without parsing the whole file.

ViewKit reads each line as a record, unions the keys across records into columns, and shows the result as a table. Nested objects and arrays render as JSON you can read in place. Malformed lines are skipped and counted rather than breaking the view, and nothing is uploaded.

What you can do with the JSON Lines viewer

How to open a JSON Lines file

1. 1

   Drop your .jsonl or .ndjson file onto the box above (or click to browse). It is read in your browser, nothing uploads.

2. 2

   Each line is parsed into a row and the keys become columns.

3. 3

   Scroll the table; nested objects and arrays are shown as JSON in the cell.

Doing the same thing in code

Reading the first few records in Python:

import json

with open("data.jsonl") as f:
    for i, line in enumerate(f):
        if i >= 5:
            break
        record = json.loads(line)
        print(sorted(record))          # the keys in this record

Finding which keys exist across the whole file requires streaming every line, collecting the union of keys, and handling lines that fail to parse. This page does that on load and reports the number of skipped lines.

JSON Lines viewer: frequently asked questions