English
Deutsch
PolskiA sitemap can be checked before it is published, without being online anywhere. The validator reads pasted XML or a file (including .gz) in the browser and checks it against the sitemaps.org protocol and against Google's extensions for hreflang, images, videos and news. Every finding names its line. Nothing is uploaded.
XML: well-formedness with line and column (unescaped &, unclosed or wrongly nested tags, unknown entities, undeclared prefixes), UTF-8 encoding, at most 50,000 entries and 50 MB uncompressed.
Protocol: root urlset or sitemapindex in the namespace http://www.sitemaps.org/schemas/sitemap/0.9, exactly one loc per entry as an absolute, encoded address under 2,048 characters, lastmod in W3C format, changefreq and priority with permitted values, order as in the XSD, duplicates, mixed hosts.
hreflang: rel="alternate", language code per ISO 639-1 with an optional region per ISO 3166-1 (Google does not accept es-419 or en-UK), self-reference in every group and return links between the entries of the file.
Images, videos, news: image:loc, at most 1,000 images per address and the image tags removed in 2022; a video's required fields, duration 1 to 28,800 s, rating 0 to 5, yes/no fields and deprecated tags; for news the publication, language, date, title, at most 1,000 entries and the two-day age limit.
With the address of the file: the sitemaps.org location rule (only addresses below the file's folder, same scheme, same host). Google goes further when the sitemap is submitted in Search Console, so a violation is shown as a note. What the addresses answer live is not checked by the validator.
How twenty large sites split their sitemaps and which violations they contain is measured in the article “Sitemap Structures Measured at 24 Sites”.
A published sitemap, with its index, child files and a sample of its addresses, is checked by the XML Sitemap Auditor.