data

JSONL Viewer

Scan JSONL / NDJSON locally and inspect invalid lines, field presence, mixed types, and records.

Local processing

About your filesFiles and content selected in this tool are processed in your browser and are not uploaded to a Browser Kitty backend.

Operation demo See the basic workflow in a short video.

About this tool

JSONL Viewer scans one-JSON-object-per-line JSONL / NDJSON files in the browser and separates valid records from lines that need review. Valid records remain available even when malformed JSON or empty lines are present.

The initial scan does not keep every record as a JavaScript object. It stores sparse checkpoints and rereads only the needed range for each page, while also reporting top-level field presence and observed JSON type mixtures.

Good for

Check large JSONL / NDJSON files for malformed JSON or empty lines
Review how often fields appear and whether types such as number, string, and null are mixed
Inspect a page as a table or formatted JSON and export only the current page to CSV

What it can do

Key capabilities available in this tool.

Scan large JSONL / NDJSON files line by line locally
Separate invalid JSON and empty lines as items to review
Inspect field presence and mixed JSON types
Table / Record views plus current-page JSONL / CSV output

How to use

Add .jsonl / .ndjson / .jsonl.txt / .ndjson.txt files.
After scanning, review total lines, valid lines, lines to review, field presence, and observed types.
Use the issue list to jump to pages containing malformed JSON or empty lines.
Switch between Table / Record views, then copy the current page as JSONL or copy/save it as CSV when needed.

Supported

Input: .jsonl / .ndjson / .jsonl.txt / .ndjson.txt.
Analysis: line-by-line Blob Worker scan, malformed JSON / empty-line separation, top-level field presence, and observed JSON types.
Viewing: multiple files, Table / Record, current-page sorting, and Cell Inspector.
Output: copy valid records on the current page as JSONL, or copy/save the current page as CSV.

Limitations and notes

.jsonl.gz / .ndjson.gz are not supported.
The source JSONL / NDJSON cannot be repaired or edited.
Sorting applies to the current page only.
JSON Schema validation, JSONPath, jq-style queries, and nested-field statistics are outside the scope; field statistics are top-level only.

Frequently asked questions

Are JSONL files uploaded to a server?

No. Selected JSONL / NDJSON files are analyzed in the browser, including the Blob Worker scan, and are not uploaded to Browser Kitty.

Does one malformed JSON line prevent the whole file from opening?

No. Malformed JSON and empty lines are counted as items to review while valid records remain viewable.

Can it convert the entire JSONL file to CSV?

CSV copy/save covers the current page only; it does not perform whole-file conversion.

Can it open both JSONL and NDJSON?

Yes. It supports .jsonl and .ndjson, plus .jsonl.txt / .ndjson.txt.

Offline version and source code

Save the single-HTML build to scan, browse, and export current-page JSONL / NDJSON data locally. The Blob Worker is generated from code inside the HTML and requires no runtime network access.

View source on GitHub