An anonymous crawl has to finish in time. Default is 500 pages; 2,000 is the hard ceiling per run. Crawl large sites in sections.

Are JavaScript routes included?

Often not. If the app is SPA-only, generate a sitemap in the app and paste it into Sitemap Validator.

Only from Last-Modified, article dates, JSON-LD dates, or a time datetime attribute. Never from the crawl clock.

Do you gzip the download?

No. You get sitemap.xml, or a zip of uncompressed XML files if the URL count hits the 50,000 protocol split.

Is robots.txt respected?

Yes for discovered links. The start URL is fetched even if robots would block it, so you can see what that page contains.

Why did example.com return one URL?

The crawl found one HTML document and no further same-host links. lastmod was 2026-08-28 from a real date, not the clock. A larger origin fills Pages versus the cap. Visual is the path tree; XML is the file you download.

What do Visual and XML mean?

Visual is the path tree for spotting missed folders. XML is the preview used for Download sitemap.xml. lastmod stays on in that preview. changefreq and priority are off until you enable them in the download menu.

Can I crawl more than 2,000 pages?

No. 2,000 is the hard ceiling so the run can finish. Default is 500. Five runs per hour. Split a large origin by section, host the files, then point an index at them. This page does not merge those files for you.

Sitemap Generator

Loading tool…

About Sitemap Generator

Enter a homepage you are allowed to crawl, set a page limit if you need to (default 500, hard ceiling 2,000), and click Crawl site. We fetch the start URL, read robots.txt, enqueue sitemap locations when they exist, then follow same-host links, five at a time. Stats show Pages, Discovered, and Blocked by robots.txt. Switch Visual and XML. Download sitemap.xml when you are done.

This is HTML crawling, not a headless browser. JavaScript-only routes are often missing. Submit the file in Search Console yourself. If a crawl would exceed 50,000 URLs, the download is a zip of a sitemap index plus sitemap-1.xml and so on. That split is a protocol safeguard; the page cap of 2,000 means a zip is unusual here.

When to use it

Use it when a site has no sitemap yet, or after an information-architecture change, and you need an XML file from a same-host HTML crawl rather than a hand-written list.

Do not use it as a validator. Download, then open Sitemap Validator on the file you host. It is not a Search Console ping.

How to use Sitemap Generator

1Enter the homepage. You should own the site or have permission to crawl it.
2Set a page limit if you need to. Default 500. Max 2,000 per run.
3Click Crawl site. Watch pages fetched versus the limit.
4Download sitemap.xml. Submit it in Search Console yourself. Use the tree tab to spot missed sections.

Dates and XML options

lastmod is included only when we found a real date: Last-Modified (dropped if it matches the Date header within two seconds, a common CDN 'now'), og:updated_time, JSON-LD dateModified, article:modified_time, JSON-LD datePublished, article:published_time, or the first time datetime attribute. Crawl time is never written as lastmod.

The download menu can add changefreq weekly and a simple priority (/ is 1.0, one path segment 0.8, else 0.5). The XML preview on screen keeps lastmod on and those extras off. Gzip is not emitted on download.

How to read the results

Stats are Pages, Discovered, and Blocked by robots.txt. Tabs are Visual (path tree) and XML. Download sitemap.xml writes the file; optional toggles add changefreq and priority. hitLimit means you reached the cap you set. timedOut means the server stopped the run.

example.com on a 25-page cap produced one loc (https://example.com/) and a lastmod of 2026-08-28 from a real date header, not the crawl clock. A JS-only app often shows a tiny tree even at 2,000.

Why this matters

A sitemap only helps if the locs are the documents a crawler can already request. lastmod invented at crawl time trains engines to refetch everything as new. That is why a CDN Date that matches Last-Modified within two seconds is dropped.

The 50,000 URL protocol split exists so one file does not break parsers. This tool's 2,000 page ceiling almost never hits that zip path. Large origins still need section crawls, then a hand-built index.

Limits and privacy

Five Crawl site runs per hour on your network. Default 500 pages, ceiling 2,000 (about 90 seconds at 500, under three minutes at 2,000). Loopback and reserved IPs are rejected. Scan only origins you may test. The crawl is dropped after download. No login.

Common mistakes

Crawling a SPA and publishing a one-URL sitemap as complete. Generate the file in the app instead.

Leaving changefreq weekly on every loc. The on-screen XML preview keeps that off for a reason.

Submitting the download without hosting it. Search Console needs a URL on your origin, not a file that only exists in this browser.

FAQs

Why a page limit?: An anonymous crawl has to finish in time. Default is 500 pages; 2,000 is the hard ceiling per run. Crawl large sites in sections.
Are JavaScript routes included?: Often not. If the app is SPA-only, generate a sitemap in the app and paste it into Sitemap Validator.
When is lastmod set?: Only from Last-Modified, article dates, JSON-LD dates, or a time datetime attribute. Never from the crawl clock.
Do you gzip the download?: No. You get sitemap.xml, or a zip of uncompressed XML files if the URL count hits the 50,000 protocol split.
Is robots.txt respected?: Yes for discovered links. The start URL is fetched even if robots would block it, so you can see what that page contains.
Why did example.com return one URL?: The crawl found one HTML document and no further same-host links. lastmod was 2026-08-28 from a real date, not the clock. A larger origin fills Pages versus the cap. Visual is the path tree; XML is the file you download.
What do Visual and XML mean?: Visual is the path tree for spotting missed folders. XML is the preview used for Download sitemap.xml. lastmod stays on in that preview. changefreq and priority are off until you enable them in the download menu.
Can I crawl more than 2,000 pages?: No. 2,000 is the hard ceiling so the run can finish. Default is 500. Five runs per hour. Split a large origin by section, host the files, then point an index at them. This page does not merge those files for you.