JSON to XML & XML to JSON Converter

1. 1. Paste or upload your XML. Switch to the XML → JSON tab, then paste XML text or upload a .xml file.
2. 2. Decide how values are typed (optional).By default, every leaf value stays a string — the safest choice for IDs and codes with leading zeros. Turn on "Convert true/false and plain numbers automatically" in Advanced options if you want ordinary booleans and numbers converted.
3. 3. Click Convert. The XML is validated first — malformed XML, an undeclared namespace prefix, or unsupported mixed content is reported clearly.
4. 4. Copy or download the resulting JSON in Pretty or Compact formatting.

An XML attribute becomes a JSON key prefixed with @, and text that sits alongside attributes or child elements becomes a key named #text. The reverse conversion understands this same convention, so JSON built this way converts cleanly back to XML attributes and text.

<product id="12">
  <name>Desk</name>
</product>

{
  "product": {
    "@id": "12",
    "name": "Desk"
  }
}

Repeated sibling elements become an array. <tag>a</tag><tag>b</tag> becomes "tag": ["a", "b"], and converting that array back produces the same two repeated elements.

Converting between JSON and XML is not universally lossless— the two formats don't model data the same way, so a few cases are worth understanding before you rely on a round trip:

• Null, empty string, and empty array all collapse — XML has no native null, so null, "", and {} all become the same empty element (<a/>), and an empty array produces no element at all. Converting that empty element back to JSON always produces null— the original distinction can't be recovered.
• Mixed content is reported, not rearranged — an element like <p>Hello <b>world</b>!</p> has text on both sides of a child element. That ordering can't be represented faithfully in this tool's JSON shape, so it's returned as a clear unsupported-content error instead of being silently reordered or dropped.
• CDATA boundaries aren't preserved — a <![CDATA[...]]> section is treated as plain text; the fact that it was originally wrapped in CDATA is not retained.
• Comments and processing instructions are omitted — neither has a JSON equivalent, so both are dropped when converting XML to JSON.
• Namespaces are preserved, not resolved — a declared namespace prefix and its xmlns attribute are kept as-is. An element using an undeclared prefix is rejected with a clear error rather than silently stripped.
• Large integers stay exact— a 19-digit database ID or a value with leading zeros is kept as text by default (and even with automatic type conversion turned on) specifically so it's never silently rounded or reformatted.