Large JSON Viewer

Why big JSON files freeze most viewers

Most tools parse on the main browser thread and build a DOM element for every node. A file with a million nodes means a million elements, which locks the page. This viewer parses in a Web Worker and renders only the few dozen rows that fit on screen, so scrolling, searching and expanding stay smooth whatever the file size. Loading the text into the editor still runs on the main thread, so very large files can pause the page briefly while they load (see the benchmarks page for measured numbers).

What to expect with huge files

On a desktop computer with Chrome a 25 MB file is parsed and shown as valid in roughly two to three seconds, and the page can pause for a second or two while the editor loads the text (measured numbers are on the benchmarks page). For very large inputs the editor switches to a lighter mode without minimap or language services. The tree view lists at most 250,000 rows at once; if you expand everything on a bigger file, collapse some branches or use search to narrow it down.

Your data stays on your device

Large files often contain real production exports, logs or customer data. Nothing is uploaded: the file is read by your browser and processed locally, so you can inspect sensitive dumps without sending them to a third party.

Why large JSON files are hard to work with

JSON is verbose. Every record repeats all of its key names, and pretty-printing adds indentation on top. A file of a few hundred thousand small records easily reaches tens of megabytes even though the information in it is modest, and the same data stored as CSV or in a binary format is often much smaller.

Memory matters as much as file size. When a program parses JSON it builds objects, arrays and strings in memory, and the result is usually several times larger than the text it came from. That is why a 25 MB file can make a tab use hundreds of megabytes, and why very large documents run into browser limits long before they run into disk limits.

Browser limits you can run into

In Chrome and other engines based on V8, a single string cannot be longer than about 512 million characters (2 to the power 29, minus 24). A JSON document is first read as one string, so files near that size cannot be opened in a browser at all, whatever tool you use. Memory limits per tab are lower and depend on the device, so phones and low-memory machines struggle with much smaller files than a desktop does.

Rendering is the other limit. Showing every node as a page element does not scale: a million nodes means a million elements and a frozen page. This viewer draws only the rows that fit on screen and asks a background thread for the rest, which is why expanding and scrolling stay smooth. Loading the raw text into the editor still takes time in proportion to its size; the benchmarks page has the measured numbers.

A workflow for big files

Start with the shape rather than the contents. Open the file, wait for the status bar to show the size, depth and node count, collapse everything, and look at the top-level keys. Most large files are one big array of similar records, or an object with one huge array inside it. Expand only the first item or two to learn what a record looks like.

Then search instead of scrolling. The search box matches keys and values and shows how many matches there are, and it opens the path to each one. When you need the same field from every record, switch to the Query tab and write a JSONPath expression, for example $.records[*].email, then copy all matches as a JSON array.

If you only need part of the file, cut it down first. Convert it to CSV with the converter if you want a spreadsheet, or generate a schema from it to understand its structure without reading it.

$.records[*].email
$.records[?(@.active == false)].id
$.records[0:10]

When a browser is the wrong tool

For files above about a hundred megabytes, or when you need to process many files, use a command-line tool. jq can filter and reshape JSON as a stream without loading it all into memory in some modes, and tools that read newline-delimited JSON handle files of any size one line at a time. A browser viewer is best for the case this page is for: looking at a large file, understanding it and extracting pieces of it.