JSONPath Query - Select Nodes from JSON in the Browser

Input - catalog JSON and path

{
  "items": [
    { "id": 1, "name": "Ada" },
    { "id": 2, "name": "Grace" }
  ]
}

JSONPath: $.items[*].id

Output - wrapped match array

[
  1,
  2
]

Collecting primary keys from a paginated catalog

List endpoints often return { items: [{ id, name }, …] } plus pagination metadata you do not care about in the moment. Path $.items[*].id is the same shape as this page's default. Paste the payload, copy the id array, and drop it into a test, a SQL IN clause, or a follow-up request. You are not writing a mapper and you are not installing jq for a thirty-second extraction.

Pulling nested error strings out of gateway envelopes

API gateways wrap failures as error.details[].message or a similar nest. The HTTP client shows a 400; the useful sentence is four objects deep. Recursive descent $..message collects every message node when you do not remember the exact envelope. If the path is stable, a precise dotted path is clearer and faster than walking the whole tree.

Harvesting SKUs or slugs before a spreadsheet import

Product dumps mix variants, images, and prices. You often need one column: SKU, handle, or slug. JSONPath returns that column as a JSON array you can paste into a sheet or feed to JSON to CSV after you rebuild a tiny array of objects. Query first when flattening the whole tree would create dozens of unused columns.

Checking which keys a generated infrastructure file actually contains

Cloud templates and policy documents are JSON with repetitive nested blocks. Before you grep a 4,000-line file, ask $..Action or $.Resources.*.Type (syntax depending on the document). You see the set of values without pretty-printing the entire template. If the file is not strict JSON, repair it first on JSON repair.

Building compact test fixtures from a production-shaped sample

A captured response is useful until it is 80 KB of noise. Selecting $.data.user.id and a handful of sibling paths gives you the values your test actually asserts. Paste those into a fixture file. The query tool does not rewrite the capture; you decide what to keep. Edit the slim document on the JSON editor afterward.

Teaching JSONPath without installing Node or jq

Workshops stall when half the room fights package managers. This page loads, shows Ada and Grace, and makes [*] versus [0] visible in seconds. Students can break the path, see an empty array or an error, and recover with Clear. That is a better first hour than a slide of XPath analogies.

Auditing which fields a client actually reads

Mobile and web clients often ignore half of a bloated payload. If you know the screens, you can write paths for the fields that appear in the UI and collect them from a real response. The leftover keys are candidates to drop from the API. Pair with JSON compare when you want a full tree diff between two backend versions rather than a list of selected nodes.

Extracting every URL nested under a CMS document

CMS JSON hides links in blocks, modules, and metadata. Recursive descent on a key such as $..href or $..url produces a checklist for link rot, CDN migration, or CSP review. It will also pick up unrelated keys named the same way - read the array before you treat it as a sitemap.

What is JSONPath?

JSONPath is a query language for JSON, analogous to XPath for XML. A path such as $.items[*].id walks from the root, into items, over every array element, and reads id. This page uses jsonpath-plus with preventEval so the path cannot run arbitrary JavaScript.

How does the default query $.items[*].id work?

The sample document is a catalog with two objects: Ada (id 1) and Grace (id 2). The default path $.items[*].id returns [1,2]. Path $.items[0].name returns ["Ada"]. Results are always wrapped as a JSON array.

Is JSONPath the same as jq?

No. jq is a programming language that can map, reduce, and define functions. JSONPath only selects nodes. There is no reduce, no custom functions, and no file globbing. Use jq on your machine when you need a full transform language.

Does this JSON query tool upload my document?

No. Matching runs entirely in the browser. The document and path stay in localStorage on this device for up to 30 days and are never sent to a server.

Why are filters such as $.items[?(@.id==1)] rejected?

Historical JSONPath engines evaluated filter scripts with JavaScript eval. That is unsafe in a web page. This tool sets preventEval and fails paths that require executing an expression. Use an index such as $.items[0] instead.

Can I transform the JSON document, not only query it?

The output is the list of matches, pretty-printed as JSON. It is not a patch engine and it does not rewrite the source. Edit the original on the JSON editor, or compare versions on JSON compare.

What does wrapping matches as an array mean?

A single match and many matches have the same shape: a JSON array. $.items[0].name returns ["Ada"], not the bare string Ada. That makes copy-paste into tests predictable.

Does recursive descent such as $..id work?

Yes. $..id is a path, not a script, so it still runs with preventEval on. It collects every id nested anywhere under the root and returns them as an array.

Can I query JSON5, JSONC, or JSON with comments?

No. The document must parse as RFC 8259 JSON first. Comments, unquoted keys, and trailing commas fail before the path runs. Repair the draft on JSON repair or validate JSON5 on the JSON5 validator, then query the strict JSON.

Is this JSONPath query tool free?

Yes. There is no registration, no usage cap, and no premium filter pack. Query as much JSON as your browser can hold in memory.