Broken Link Checker: Find 404s, Dead Links, Redirect Chains & SEO Issues

Broken link checker and dead link finder for any list of URLs, a website page, or a sitemap.xml. Each link is checked with a lightweight HTTP request. You get the status code, final URL after redirects, redirect chain, response time and a clear classification (ok, not_found, redirect, server_error, timeout, ssl_error, …).

No login, no API key, no headless browser and no paid proxy. Built for SEO audits, site migrations, content QA and scheduled health checks.

Why this broken link checker

	This Actor	Typical checkers
Input	URL list, start pages (<a href> scan) or sitemap	Often just a URL list
Redirect chains	Full chain + hop count	Final status only
HEAD with GET fallback	✅ sites that block HEAD still work	HEAD-only misses many URLs
robots.txt on page scans	✅ respected by default	Rare
Return only broken rows	✅ so you don't pay for healthy links	Often charge every URL
Price	$0.90 per 1,000 links checked	$1.00–$1.20 / 1,000, sometimes plus a $0.02 start fee

A popular Store checker in this niche sits around 2.7★ with reports of missed redirects and noisy results. This Actor records the full redirect chain, classifies SSL and timeout errors separately, and lets you export only the rows that need a fix.

Who uses it

• SEO agencies: crawl a client's sitemap or key landing pages and deliver a 404 / redirect report.
• Site migrations: confirm old URLs 301 to the right new paths and that nothing 404s.
• Content and docs teams: scan published articles for dead outbound citations.
• Developers: schedule a daily run and alert Slack when new not_found or server_error rows appear.

Example input

Check a mix of known URLs:

{
  "urls": [
    "https://www.python.org",
    "https://www.python.org/this-page-does-not-exist"
  ],
  "onlyBroken": false,
  "maxLinks": 50
}

Scan a page for links, then check them:

{
  "startPages": ["https://www.python.org"],
  "checkSameDomainOnly": false,
  "includeResourceLinks": false,
  "onlyBroken": true,
  "respectRobotsTxt": true,
  "maxLinks": 100
}

Or paste sitemap URLs to verify every <loc>.

Example output

{
  "url": "https://www.python.org/this-page-does-not-exist",
  "ok": false,
  "classification": "not_found",
  "status": 404,
  "finalUrl": "https://www.python.org/this-page-does-not-exist",
  "redirectChain": [],
  "redirects": 0,
  "contentType": "text/html; charset=utf-8",
  "responseMs": 187,
  "error": null,
  "sourcePage": null,
  "anchorText": null,
  "origin": "direct",
  "checkedAt": "2026-11-06T07:35:00+00:00"
}

A SUMMARY record in the default key-value store counts links planned vs written, broken down by classification.

Pricing

Pay per event:

Event	Price
Actor start	$0.00005
Link checked (row written)	$0.0009 ($0.90 per 1,000)

Turn on Return only broken or redirected links to skip writing (and paying for) healthy 200 responses. Set a maximum charge per run to cap spend.

API example

curl -X POST "https://api.apify.com/v2/acts/classier_wholemeal~broken-link-checker/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"urls": ["https://www.python.org", "https://www.python.org/missing"], "onlyBroken": true}'

Export as JSON, CSV or Excel, or connect Google Sheets, Zapier, Make, n8n, Slack or webhooks. Works as a tool for AI agents through the Apify MCP server.

FAQ

Will this crawl an entire website? It checks the URLs you give it, every link on the start pages you list, and/or every URL in the sitemaps you paste. It does not spider the whole site recursively, which keeps runs predictable and polite.

Does it execute JavaScript? No. Links injected only by client-side JS will not be discovered from start pages. Sitemap and explicit URL inputs do not have that limit.

Are mailto and tel links checked? No. They are skipped. So are javascript: and in-page # anchors.

Is this allowed? The Actor sends ordinary HTTP HEAD/GET requests the same way a browser or uptime monitor would, identifies itself with a clear User-Agent, and honors robots.txt when it fetches start pages. You are responsible for checking the sites you target and for how you use the report.

Why did a working page come back 403 or timeout? Some hosts treat automated clients differently. The Actor retries transient 429/5xx errors, but it will not rotate proxies or bypass bot protection.

How is a redirect classified? Any URL that returns 3xx (or a chain of them) is redirect. The full redirectChain is included. Turn on Treat 3xx redirects as issues if you want those rows when Return only broken is enabled.