Broken Link Checker & Bulk URL Status Checker
Pricing
from $0.80 / 1,000 checked links
Broken Link Checker & Bulk URL Status Checker
Find broken links, images and scripts on any website. Crawls the site, checks every URL once and shows which pages contain each dead link. Also checks plain lists of URLs. HTML report, CSV and JSON output.
Pricing
from $0.80 / 1,000 checked links
Rating
0.0
(0)
Developer
John
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
30 minutes ago
Last modified
Categories
Features
Share
Find every broken link, image, script and stylesheet on a website. The Actor crawls the site, checks each unique URL once and tells you which links are broken and which pages contain them, so you know exactly where to fix them.
It can also check only the links on pages you list, or a plain list of URLs, without crawling.
Who it is for
- Site owners and developers who want to catch 404s, dead external links and missing images before visitors do, or after a migration or redesign.
- SEO agencies and freelancers who need a clean broken-link report for a client, with source pages and anchor text.
- Content teams who schedule a weekly check and fix what turns up.
- AI agents and automations that need link status as structured data (JSON, CSV, Excel).
What you get
- One dataset row per checked URL, broken ones first: result, status code, a plain-English reason, link type, how many pages link to it, the pages themselves, anchor text, redirect chain and final URL.
- An HTML report (
REPORTin the key-value store) you can open in a browser or send to a client: broken links, the pages with the most broken links, links to check manually, and redirected links. - A CSV of broken links by page (
BROKEN_LINKSin the key-value store): one line per broken link and page that contains it, ready for a developer or a spreadsheet. - A JSON summary (
OUTPUTin the key-value store) with counts by result and status code, top broken links and what was not checked.
How results are classified
| Result | Meaning |
|---|---|
| broken | The link does not work for visitors: 404, 410 and other 4xx errors, 5xx errors, a domain that does not resolve, a refused connection, a TLS certificate error, a redirect loop, or a redirect that ends in an error. |
| uncertain | The server refused or did not answer an automated request: 401, 403, 429, a timeout or a dropped connection. These links often work in a browser, so they are listed as "check manually" and never counted as broken. |
| redirect | The link works but goes through one or more redirects. The row shows every hop and the final URL. |
| ok | The link works. |
Each URL is checked with a lightweight HEAD request first. If the server answers HEAD with an error, the Actor tries a normal GET before reporting a problem, which avoids the false alarms that HEAD-only checkers produce.
Three ways to use it
| Mode | What it does |
|---|---|
| Crawl the site (default) | Starts at your URL, follows internal links and checks every link and resource found on every page. |
| Check the links on the pages I list | Reads only the pages you give it and checks the links on them. Good for landing pages or a single article. |
| Check only the URLs I list | A bulk URL status check: no pages are read, each URL is checked once. |
How to use it
- Enter one or more Start URLs, usually the homepage.
- Set Maximum links to check. Start with 1,000 to get a feel for the output.
- Run the Actor. Open the Output tab for the table, or the key-value store for the
REPORT,BROKEN_LINKSandOUTPUTrecords.
Input options
| Field | What it does | Default |
|---|---|---|
startUrls | The site to crawl, or the pages or URLs to check | required |
mode | crawl, pages (links on the listed pages) or urls (the listed URLs only) | crawl |
maxLinks | Stop adding new links after this many unique URLs | 1000 |
checkExternalLinks | Check links that point to other websites | true |
checkResources | Check images, scripts, stylesheets, icons, frames and media files | true |
crawlScope | hostname, domain (include subdomains) or path (stay in the start folder) | hostname |
useSitemap | Also check the URLs listed in XML sitemaps | false |
respectRobotsTxt | Respect robots.txt and Crawl-delay on the crawled site | true |
maxCrawlPages | Stop reading new pages for links after this many (0 = no separate limit) | 0 |
maxDepth | Maximum clicks from a start URL | 20 |
maxConcurrency | Requests in parallel across all hosts | 10 |
excludeUrlPatterns | Skip links whose URL matches these texts or * wildcards | none |
userAgent | Custom User-Agent string | SiteAuditBot |
Example input
{"startUrls": [{ "url": "https://www.example.com/" }],"maxLinks": 5000,"excludeUrlPatterns": ["*/calendar/*", "?replytocom="]}
Example output (one row)
{"url": "https://www.example.com/old-pricing","result": "broken","isBroken": true,"statusCode": 404,"reason": "404 Not Found","linkType": "internal","resourceType": "link","finalUrl": null,"redirectCount": 0,"redirectChain": null,"foundOnCount": 12,"foundOn": ["https://www.example.com/", "https://www.example.com/blog/launch"],"anchorTexts": ["See pricing", "Plans"],"contentType": "text/html","responseTimeMs": 143,"method": "GET","depth": 1,"crawled": false,"linksOnPage": null,"inSitemap": false,"checkedAt": "2026-11-17T12:00:00.000Z"}
Pricing
You pay per result: $1 per 1,000 links checked, plus Apify's small Actor start fee. Customers on Apify's Starter, Scale and Business plans pay $0.95, $0.90 and $0.80 per 1,000. There is no monthly rental and you do not pay for platform compute separately.
Every unique URL that is checked produces one result: pages, external links, images and other resources alike. A URL linked from 500 pages is still checked and charged once. Links that are not checked, for example because another site's robots.txt disallows it, are not charged.
Set Maximum links to check, or a maximum cost per run in Apify Console, to cap your spend. The Actor never checks more links than your limit can pay for.
| Links checked | Approximate cost |
|---|---|
| 1,000 | $1 |
| 10,000 | $10 |
| 100,000 | $100 |
Good to know
- HTML is read as the server sends it. Links that JavaScript adds after the page loads are not seen, so single-page apps without server-side rendering will show few links.
- The crawler is polite. It identifies itself as
SiteAuditBot, sends at most about 10 requests per second to the site it crawls and at most 2 per second to any other host, and honoursCrawl-delayup to 3 seconds. - robots.txt is respected. On the site you crawl you can turn this off if you own the site. The robots.txt files of other websites are always respected: links they disallow are listed as not checked.
- If your firewall blocks it, allowlist the
SiteAuditBotuser agent or set a custom User-Agent that your firewall accepts. - What is checked:
a,area,img,script,link(stylesheets and icons),iframe,video,audio,source,track,embedandobjectURLs. Links insidenoscriptare included. Files referenced only from CSS, andsrcsetvariants, are not checked. - Not checked:
mailto:,tel:andjavascript:links, and whether a#fragmentexists on the target page. - Local addresses: links from a public site to
localhostor a private network address are flagged for a manual check without being requested. - Limits: up to 100,000 links per run and 5 MB of HTML per page.
- Run timeout: a run stops at the timeout set in its run options (one hour by default) and saves everything checked up to that point. As a guide, one site is checked at up to 10 URLs per second, so allow about an hour for every 30,000 links and raise the timeout for larger or slower sites. A run stopped with Abort gracefully also saves what it has checked so far; Abort now stops at once and saves nothing.
- Memory: pages are read without building a document tree, so the default 1 GB is enough even for sites with very large pages.
- This is an independent tool. It is not affiliated with or endorsed by any search engine.
FAQ
The run finished and mentions broken links in its status. Did something go wrong? No. Those numbers are the links found on the website you checked. A run that completed shows the status Succeeded.
Why is a link listed as "check manually" instead of broken? The server answered with 401, 403 or 429, or did not answer in time. Many sites refuse automated requests but work in a browser, so reporting them as broken would be wrong. Open them yourself to confirm.
Why was a link not checked at all? Either the target site's robots.txt disallows automated requests to it, the maximum number of links was reached, or you excluded it. The report lists the counts.
How do I see which page a broken link is on?
Every row has foundOn with up to 20 pages, and the BROKEN_LINKS CSV has one line per broken link and page.
Can I check a list of URLs without crawling? Yes. Put the URLs in Start URLs and choose "Check only the URLs I list". In that mode the link type, source pages and anchor text stay empty, because there is no site to relate them to.
Can I schedule it? Yes. Use Apify schedules to run a weekly or monthly check and compare the summaries.
Related tools
- Technical SEO Audit & Site Crawler: crawls a site and audits every URL with 57 technical and on-page SEO checks, a score per page and an HTML report.
- Website Technology & Tech Stack Detector: finds the CMS, e-commerce platform, analytics, frameworks and other technologies that a list of websites use.
Feedback
Found a false positive or want something added? Open an issue on the Issues tab and it will be looked at.