Free JSONPath & XPath Tester

How this tool computes its result

JSONPath mode parses the input with `JSON.parse` and evaluates the expression with the `jsonpath-plus` library's `JSONPath({ path, json })`, returning an array of matches. XPath mode parses the input as XML with the browser's `DOMParser` (throwing "Invalid XML input." if a `parsererror` node appears), then runs `document.evaluate(expression, doc, null, XPathResult.ANY_TYPE, null)`. Depending on the result type returned by the browser, it pushes a single number, string, or boolean, or iterates node results with `iterateNext()` collecting each node's `textContent`.

Worked example

JSONPath mode: input `{"items":[{"name":"Brand Armor AI"},{"name":"Acme"}]}` with expression `$.items[*].name` returns 2 matches: `["Brand Armor AI","Acme"]`. XPath mode: input `<items><item><name>Brand Armor AI</name></item></items>` with expression `//item/name/text()` returns 1 match: `["Brand Armor AI"]`.

When not to use this tool

XPath evaluation depends on the browser's XML parser and `document.evaluate`, so it only works on well-formed XML, not arbitrary HTML — tag soup that a browser would silently repair in an HTML context throws a parse error here. It is also not a JSON Schema or XML Schema validator; it only reports what an expression matches, not whether the document is structurally valid.

Common mistakes

• - Leaving JSON content in the input box after switching to XPath mode (or vice versa) — the wrong parser runs against it and throws immediately rather than auto-detecting the format.
• - Writing an XPath expression without a trailing `text()` — it returns Element nodes whose `textContent` can include nested whitespace instead of the clean string value.
• - Referencing a top-level array in JSONPath without the leading `$.` — this silently returns zero matches rather than an error, which can look like the data is missing.

Related tools