Sitemap URL Extractor — get all URLs of any website, and see what changed

Paste a website or a sitemap URL and get a clean, deduplicated list of every URL the site publishes in its sitemap.xml — with lastmod, changefreq, priority, images, videos, news and hreflang alternates. Export it to CSV, JSON or Excel with one click. Nested sitemap indexes, gzipped .xml.gz files and plain-text sitemaps are handled automatically, and the sitemap is discovered for you through robots.txt when you only give a domain.

Typical questions it answers: list all pages of a website, extract all URLs from a sitemap, which pages did a site add or remove this week, which sitemap URLs return 404.

Two things set it apart from other sitemap tools on Apify Store:

• It streams. Sitemaps are parsed as they download, so memory stays flat whether a site has 500 URLs or 5 million. No timeouts on huge indexes, no out-of-memory crashes.
• Delta mode. Turn it on and each run reports which URLs were added or removed since your previous run of the same site. That turns a URL list into a content-change monitor for SEO, competitor tracking or crawl scheduling — without you storing anything.

It sends no requests to the pages themselves unless you ask for HTTP status checks, and it needs no proxy: sitemaps are public files written for crawlers.

Output: one row per URL, exportable to CSV, JSON, Excel

One dataset row per URL (download the whole list from the Export button, or fetch it via API):

{
  "url": "https://example.com/blog/streaming-xml",
  "lastmod": "2026-09-12T08:14:00+00:00",
  "changefreq": "weekly",
  "priority": 0.8,
  "sitemap": "https://example.com/sitemap-posts.xml",
  "site": "https://example.com",
  "depth": 1,
  "images": ["https://example.com/img/hero.png"],
  "videos": 0,
  "newsTitle": null,
  "alternates": [{ "hreflang": "de", "href": "https://example.com/de/blog/streaming-xml" }],
  "change": "added",
  "statusCode": 200,
  "finalUrl": null
}

Field	Meaning
url	The page URL, exactly as listed in the sitemap. Deduplicated across all sitemap files of the site.
lastmod, changefreq, priority	Values declared in the sitemap, or null when absent.
sitemap	The sitemap file the URL came from. null for URLs that were removed since the last run.
site	Origin of the start URL.
depth	Nesting level of the sitemap file: 0 for the top-level sitemap, 1 inside a sitemap index, and so on. -1 for removed URLs.
images, videos, newsTitle, alternates	Data from the image, video, news and xhtml (hreflang) sitemap extensions. Turn off Include image, video, news and hreflang data if you do not need them.
change	Only in delta mode: added, removed or unchanged.
statusCode, finalUrl, statusError	Only with Check HTTP status: the response code, the URL after redirects (when different) and the error if the request failed.

A run summary — sitemaps fetched, URLs found, filtered and pushed, errors, delta counts and timing — is saved to the SUMMARY record of the run's key-value store.

Input

Option	Default	What it does
Websites or sitemap URLs	—	One or more start URLs. A bare domain is resolved via robots.txt (Sitemap: lines) and then /sitemap.xml, /sitemap_index.xml, /wp-sitemap.xml and other common locations. Anything that already looks like a sitemap (.xml, .xml.gz, .txt, a feed) is fetched directly.
Maximum URLs	0 (no limit)	Hard cap across all sites. The cap is exact, even with parallel downloads.
Include only URLs matching / Exclude URLs matching	—	Case-insensitive regular expressions, e.g. include /blog/, exclude \?page=.
Modified since	—	Keep only URLs whose lastmod is on or after an ISO date (2026-01-31) or a relative window (7d, 48h, 2w). URLs without lastmod are dropped when this is set.
Delta mode	off	Compare with the previous run of the same site and flag every URL as added, removed or unchanged. See below.
Delta output	changes only	Return only added and removed URLs, or all URLs with their flag.
Snapshot store name	sitemap-url-extractor-snapshots	Named key-value store that keeps one snapshot per site between runs.
Check HTTP status	off	One HEAD request per URL (GET fallback for servers that refuse HEAD). Adds statusCode and finalUrl. Billed separately, see Pricing.
Advanced		Status-check concurrency, sitemap download concurrency, index nesting depth, maximum sitemap files per site, custom User-Agent, optional proxy.

Sitemap change monitor: new and removed pages since the last run

Every site you extract in delta mode gets a snapshot — a compressed list of its URLs — in the named key-value store. On the next run the Actor compares the fresh list with the snapshot:

• URLs present now but not before → added
• URLs present before but not now → removed (they are still emitted, with sitemap: null and depth: -1)
• everything else → unchanged (emitted only when Delta output is All URLs)

The first run of a site returns everything as added. Schedule the Actor daily or weekly and the dataset of each run is the change log.

Two safeguards keep the delta honest:

• A truncated walk never overwrites the snapshot. If the run stops early because of Maximum URLs or because your maximum charge was reached, the snapshot is left untouched — otherwise the next run would report thousands of false removed URLs.
• Filters are part of the snapshot. If you change the include/exclude regex between runs, the difference shows up as added/removed URLs, which is what you would expect. Use different Snapshot store name values to track the same site under different filters.

Snapshots are stored as SHA-1 digests and gzipped shards of 200,000 URLs; a site with one million URLs uses roughly 100 MB of memory during the comparison.

Pricing

Pay per event, no subscription:

• one event per URL row written to the dataset (added, removed and unchanged rows alike);
• one event per URL checked when Check HTTP status is on.

Set Maximum URLs or a maximum charge on the run to cap the cost of very large sites. When the maximum charge is reached the Actor stops cleanly, keeps everything it already extracted, and — in delta mode — does not touch the snapshot.

Why no charge for platform usage: parsing is streamed and the Actor does not visit the pages, so a million-URL site costs a few cents of compute. You pay for results.

Use cases

• SEO audits — inventory every indexable URL, compare with your crawl or with Google Search Console, find orphaned or missing pages.
• Content change monitoring — schedule delta mode against your own site or a competitor's and get the new and removed URLs each morning.
• Crawl scheduling — feed the URL list (optionally filtered by modifiedSince) into a scraper so it only visits what changed.
• Migrations — export the full list before a redesign, run status checks after, and diff.
• Data pipelines and AI agents — deterministic JSON per URL, exact caps, no proxies, no login: a safe tool to call programmatically.

Run it from code

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("bruco3/sitemap-url-extractor").call(run_input={
    "startUrls": [{"url": "https://example.com"}],
    "deltaMode": True,
    "includeRegex": r"/blog/",
})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["change"], item["url"])

curl -X POST "https://api.apify.com/v2/acts/bruco3~sitemap-url-extractor/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"startUrls":[{"url":"https://example.com/sitemap.xml"}],"maxUrls":1000}'

Works with the Apify API, scheduler, webhooks and integrations (Make, Zapier, n8n, Google Sheets).

FAQ

Do I need a proxy? No. Sitemaps are public files intended for search-engine crawlers. A proxy option exists for the rare site that blocks datacenter IPs.

Does it visit the pages? Not unless you enable Check HTTP status. Extraction reads only the sitemap files.

What about huge sites? The parser is incremental: a 300 MB sitemap index is processed in a streaming fashion with constant memory. Millions of URLs are fine; use Maximum URLs if you only want a sample.

Gzipped sitemaps? Detected by content, not by file extension — .xml.gz files and gzip served without the extension both work.

Plain-text sitemaps and RSS/Atom feeds? Supported, as allowed by the sitemaps.org protocol. Feed items yield url and lastmod.

The site has no sitemap. The run reports it in the summary and fails only if no site produced any URL. There is deliberately no HTML crawling fallback: that is a different tool with different costs.

Is it legal? Sitemaps are published by site owners specifically for automated consumption. The Actor identifies itself with a clear User-Agent, downloads only sitemap files by default, and stores no personal data.

Limitations

• URLs are taken as listed; the Actor does not normalise or canonicalise them beyond deduplication.
• removed detection requires a complete walk of the site; runs cut short by a cap or budget skip it and preserve the snapshot.
• Sitemaps behind authentication or bot walls are out of scope by design.

Related

• Broken Link Checker — checks every URL of these sitemaps for 404s, redirect chains, SSL errors and slow pages, with new / persisting / fixed between runs.

Feedback

Found a sitemap this Actor cannot read, or missing a field you need? Open an issue on the Issues tab — sitemap edge cases get fixed quickly, and each one makes the tool better for everyone. If it saved you time, a review helps other people find it.

Changelog

• 0.1 — initial release: streaming parser, nested indexes, gzip, text and feed sitemaps, robots.txt discovery, regex and lastmod filters, delta mode, optional status checks, exact caps, pay-per-event.