How do I pretty print XML?

Paste the XML or press Load File, choose 2 spaces, 4 spaces or Tabs, and press Format. Copy takes the result, and Download .xml saves it as formatted.xml, or minified.xml after Minify. The output belongs to the text it was made from: after an edit or a new indent the page asks you to press Format again. The Tree view shows the same result as collapsible elements with a search box.

Does the XML formatter change my text or attribute values?

No. Attribute values keep their quotes and characters, so y='2' stays in single quotes, and text is never trimmed: padded keeps both pairs of spaces. What changes is the white space between elements and inside a tag: minifies to . A no-break space is text, not white space, so it stays.

Why is part of my XML not indented?

An element that mixes text and child elements, such as Hello world again , is copied on one line, because adding line breaks there would change the text. The same goes for anything under xml:space="preserve" and for elements that hold only text.

Does it validate XML against a DTD or schema?

No. It checks that the document is well-formed with the browser's own XML parser and shows the reason with the line and column, for example a closing tag that does not match, half a second after you stop typing. It does not validate against a DTD or an XSD, and it does not expand entities.

How large an XML file can it format?

Up to 300,000 characters; the Limits table above lists the file size, nesting, highlighting and tree limits. For a larger file, such as a full database export, the page suggests a command-line tool: xmllint --format file.xml prints the same kind of indented output.

Is my XML saved or uploaded?

A loaded file is read in your browser and is not uploaded. The input, the indent and the view are saved in this browser's local storage a moment after you stop typing, until you press Clear Saved Values.

XML Formatter

Format or minify XML with 2 spaces, 4 spaces or tabs. Comments, CDATA, DOCTYPE, attribute values and mixed content stay as written, up to 300,000 characters. How it works

Input XML

Indent

Up to 300,000 characters (files up to 2 MB). Only the spaces, tabs and line breaks between elements change (any other character, such as a no-break space, is text and stays).

Output

Formatted XML will appear here...

What changes and what does not

Comments, processing instructions (also after the root), the whole DOCTYPE, CDATA sections, entity references, attribute values and text are copied as written. Mixed content (text next to child elements), text-only elements and anything under xml:space="preserve" are never re-flowed or trimmed, because their spaces are data.

The tool checks well-formedness with the browser's XML parser. It does not validate against a DTD or schema. The XML declaration is kept as written; the text you copy is Unicode whatever its encoding= says.

About the XML Formatter

Format puts each child of an element-only element on its own line, indented by 2 spaces, 4 spaces or a tab, so <a>1<c/></a> becomes four lines; Minify removes that white space again. Only the spaces, tabs and line breaks between elements change: comments, CDATA, the DOCTYPE, attribute values, mixed content and anything under xml:space="preserve" are copied as written.

What Format and Minify do

Input	Format (2 spaces)	Minify
<a><b>1</b><c/></a>	<a> ⏎ <b>1</b> ⏎ <c/> ⏎ </a>	<a><b>1</b><c/></a>
<p>Hello <b>world</b> again</p>	<p>Hello <b>world</b> again</p>	<p>Hello <b>world</b> again</p>
<a xml:space="preserve"> <b> x </b> </a>	<a xml:space="preserve"> <b> x </b> </a>	<a xml:space="preserve"> <b> x </b> </a>
<a><!-- note --><![CDATA[ <raw> ]]></a>	<a><!-- note --><![CDATA[ <raw> ]]></a>	<a><!-- note --><![CDATA[ <raw> ]]></a>
<a x="1" y='2' ><b /></a>	<a x="1" y='2'> ⏎ <b/> ⏎ </a>	<a x="1" y='2'><b/></a>
<a><b> padded </b></a>	<a> ⏎ <b> padded </b> ⏎ </a>	<a><b> padded </b></a>

⏎ marks a line break in the output. Mixed content (text next to a child element), xml:space="preserve" and an element holding a comment next to CDATA are copied on one line as written; text inside an element is never trimmed.

Limits

What	Limit
Input	300,000 characters (files up to 2 MB)
Nesting for Format, Minify and the tree	1,000 levels
Highlighted output	50,000 characters; longer output is shown as plain text
Tree view	20,000 nodes, 100 children of a branch at a time

Common well-formedness errors and the message in Chrome

Input	What is wrong	Message shown
<a><b>1</c></a>	The closing tag does not match the open element	Invalid XML at line 1, column 12: Opening and ending tag mismatch: b line 1 and c
<a><b>1</b>	An element is never closed	Invalid XML at line 1, column 12: Premature end of data in tag a line 1
<a x=1/>	An attribute value without quotes	Invalid XML at line 1, column 6: AttValue: " or ' expected
<a>Fish & chips</a>	A bare & in text; write &	Invalid XML at line 1, column 10: xmlParseEntityRef: no name
<a>1 < 2</a>	A bare < in text; write <	Invalid XML at line 1, column 7: StartTag: invalid element name
<a/><b/>	Two root elements	Invalid XML at line 1, column 5: Extra content at the end of the document
<a> </a>	An HTML entity that XML does not define; write	Invalid XML at line 1, column 10: Entity 'nbsp' not defined

Messages as the page showed them in Chrome 152 on Linux on 22 October 2026. The text after the position comes from the browser’s XML parser, so Firefox and Safari word it differently and can point at another column.

Frequently asked questions

How do I pretty print XML?: Paste the XML or press Load File, choose 2 spaces, 4 spaces or Tabs, and press Format. Copy takes the result, and Download .xml saves it as formatted.xml, or minified.xml after Minify. The output belongs to the text it was made from: after an edit or a new indent the page asks you to press Format again. The Tree view shows the same result as collapsible elements with a search box.
Does the XML formatter change my text or attribute values?: No. Attribute values keep their quotes and characters, so y='2' stays in single quotes, and text is never trimmed: padded keeps both pairs of spaces. What changes is the white space between elements and inside a tag: <a x="1" y='2' ></a> minifies to <a x="1" y='2'></a>. A no-break space is text, not white space, so it stays.
Why is part of my XML not indented?: An element that mixes text and child elements, such as Hello world again, is copied on one line, because adding line breaks there would change the text. The same goes for anything under xml:space="preserve" and for elements that hold only text.
Does it validate XML against a DTD or schema?: No. It checks that the document is well-formed with the browser's own XML parser and shows the reason with the line and column, for example a closing tag that does not match, half a second after you stop typing. It does not validate against a DTD or an XSD, and it does not expand entities.
How large an XML file can it format?: Up to 300,000 characters; the Limits table above lists the file size, nesting, highlighting and tree limits. For a larger file, such as a full database export, the page suggests a command-line tool: xmllint --format file.xml prints the same kind of indented output.
Is my XML saved or uploaded?: A loaded file is read in your browser and is not uploaded. The input, the indent and the view are saved in this browser's local storage a moment after you stop typing, until you press Clear Saved Values.

Reading a one-line SOAP response or sitemap

Format a SOAP response, an Android layout, a Maven pom.xml, an RSS feed or a sitemap.xml that arrived on one line, so you can read it; minify a payload before pasting it into a request; or find the line and column where a document stops being well-formed. Load Sample fills in a SOAP envelope with a CDATA description, and Load File reads a .xml file of up to 2 MB in your browser and does not upload it. Download .xml saves the result as formatted.xml, or minified.xml after Minify.

How Format and Minify work

The formatter works on the text, token by token, not on a rebuilt DOM. In element-only content every child goes on its own line, indented depth × unit (2 spaces, 4 spaces or a tab), and text made only of spaces, tabs, CR and LF between elements is replaced by that line break and indent (Format) or removed (Minify). Everything else is copied as written: the XML declaration, the DOCTYPE with its internal subset, comments and processing instructions, CDATA, entity references, attribute values with their quotes, and all text. Mixed content (text next to child elements), text-only elements and anything under xml:space="preserve" stay exactly as typed. Inside a tag, the space between attributes becomes one space and  is written .

Example: <a>1<c/></a> formats to four lines, <a>, 1, <c/> and </a>, and minifies back to <a>1<c/></a>. The browser's XML parser also checks that the document is well-formed, half a second after you stop typing and again before Format or Minify, and reports the line and column of the first error.

Fun Fact: XML 1.0 became a W3C Recommendation on 10 February 1998. Its editors were Tim Bray, Jean Paoli and C. M. Sperberg-McQueen.

Related Tools

View all tools

Dev

CSS Minifier

Paste CSS to remove comments and unneeded whitespace while keeping descendant spaces, calc() operators and /*! */ license comments, with the characters saved.

Dev

JS Minifier

Paste JavaScript to remove comments and unneeded whitespace. Line breaks the code relies on and /*! license comments stay; variable names are not shortened.

Dev

HTML Minifier

Paste HTML to drop comments and collapse whitespace. The space between inline elements stays, pre and textarea are kept, and inline CSS and JS can be minified.

Dev

SQL Query Formatter

Format SQL for Standard SQL, PostgreSQL, MySQL, SQLite or SQL Server, with keywords kept, upper-cased or lower-cased and 2-space, 4-space or tab indents.

Dev

JSON Formatter

Paste JSON to format it with 2 spaces, 4 spaces or tabs, minify it or sort its keys. A syntax error is reported with its line, column and a caret under it.

Dev

CSV to XML

Convert CSV to XML: each row becomes a row element with one child per column, headers become valid XML names, and quoted commas and line breaks are kept.