On this page
Quick picks from the best free broken link checkers #
- Small site audit: Screaming Frog SEO Spider. Trace a failed URL back to its linking pages in the desktop report.
- Browser report: Dr. Link Check. Free Lite covers two projects and 1,500 links per project.
- Occasional manual scan: Dead Link Checker. Start from a public website without installing software.
- Larger free page allowance: BrokenLinkCheck.com. Its free limit is 3,000 pages.
- A site you can verify: Ahrefs Webmaster Tools. Free Site Audit credits support ongoing owner checks.
- Links and anchors in one document: W3C Link Checker. A focused public checker with optional recursion.
- Broader local report: SiteOne Crawler. Open-source crawling with HTML and machine-readable reports.
- Docs in a repository: lychee. Check links in Markdown and built HTML during development.
Compare the free allowances carefully #
| Tool | Best for | Platforms | Free limits | Paid option | Where it runs | Standout strength |
|---|---|---|---|---|---|---|
| Screaming Frog | Finding source pages for repairs | Windows, macOS, Linux | 500 URLs per crawl | Annual license adds larger crawls and advanced features | Your desktop | Inlinks and exports |
| Dr. Link Check | Browser-based project reports | Web | Two projects, 1,500 unique links each; one-time checks | Paid plans add allowances and scheduled checks | Service servers | Simple project-based scans |
| Dead Link Checker | Manual checks of public sites | Web | Manual checks free; full scan depth limited to 10 | Auto Check subscriptions | Service servers | Single or multiple-site checking |
| BrokenLinkCheck.com | Checking more pages online | Web | 3,000 pages; hyperlinks within them aren't capped | Commercial version for larger jobs | Service servers | Link location in source HTML |
| Ahrefs Webmaster Tools | Auditing a verified property | Web | 5,000 crawl credits per project per month | Broader Ahrefs subscriptions | Cloud crawler | Owner audit alongside SEO data |
| W3C Link Checker | A page's links and anchors | Web | Public checker; choose recursion depth | No paid account required | Service servers | Anchors and referenced objects |
| SiteOne Crawler | Wider technical reports | Windows, macOS, Linux | Free MIT project; hardware and crawl settings apply | No license subscription required | Local application | HTML, JSON and text output |
| lychee | Source docs and built files | Windows, macOS, Linux | Free; remote URL input checks that page, without recursive site crawling | No paid plan | CLI on your computer or build runner | Markdown and HTML inputs |
Official references: Screaming Frog, Dr. Link Check plans, Dead Link Checker FAQ, BrokenLinkCheck.com, Ahrefs free allowances, W3C checker, SiteOne repository and lychee usage.
Screaming Frog: the best starting point for a small repair list
The detail I'd look for first is the lower Inlinks panel. Selecting a failed destination shows the pages linking to it, so you can go straight to the content that needs changing. The official broken-link tutorial explains this report and its bulk export.
Users in a September 2026 discussion praised the crawler's detailed reporting and configuration. One reply also acknowledged that the interface looks less polished than a broad SEO suite. For a repair job, I'd prioritize a clear source-and-destination report over a pretty overview.
The free version crawls 500 URLs on Windows, Mac or Linux. That's its serious restriction, but it still wins for a small site because the report helps you act on the findings. Larger crawls and advanced features require an annual license.
Don't count only your published articles when judging that allowance. A site can expose many more URLs than the number of posts in its editor. Start with a modest audit and check whether the crawler reached the end.
Pick it if you want a practical desktop repair report for a small website. Skip it if a full site scan needs more than the free URL allowance. Use the Screaming Frog guide for the basic workflow.
Dr. Link Check: a browser report with a link-based allowance
Dr. Link Check is my browser pick when you want a project report instead of a desktop application. Its free Lite plan includes two projects with 1,500 unique links each and one-time checks.
The interesting limit is what counts as a link. HTML documents, images, stylesheets, scripts and other referenced resources count. Repeated references to the same resource count once. A 100-page website can therefore use a much larger share of the allowance than its page count suggests.
I'd consider it for a small public site that changes occasionally. You can review the report and work through the content edits without keeping a desktop crawler installed. For someone responsible for a company site alongside other work, that smaller setup is appealing.
The drawback is recurring monitoring: scheduled scans belong to paid plans. You'll need to remember to repeat the free check after a redesign or a round of content changes. Paid plans also increase the number of projects and links.
Pick it if you want occasional browser-based scans of one or two public sites. Skip it if you need free scheduled monitoring or your resource count regularly exceeds Lite.
Dead Link Checker: manual scans without an installation
Dead Link Checker offers a free interactive scan of a public website or individual page. Its homepage also offers a multi-site route, while automatic checks and emailed recurring reports belong to subscriptions.
I'd use it when a site owner asks a small, immediate question: are the links in this section still working? It avoids the setup of a desktop audit tool. The report can become a straightforward editing list for someone who knows where the content lives.
The FAQ explains several restrictions. It can't scan password-protected pages or an intranet. Its full scan follows links to a depth of 10, and it doesn't execute client-side JavaScript to discover additional content. That rules it out as a complete audit of a private app or a heavily scripted navigation system.
Free manual checks are available, including multiple-site checks after signing in with an email address. Paid Auto Check is the route for regular scheduling. I'd make that choice according to whether someone will actually return to the free report.
Pick it if you need an occasional scan of ordinary public pages. Skip it if the important content requires login or JavaScript-generated navigation.
BrokenLinkCheck.com: more pages, with source locations
BrokenLinkCheck.com gives its free scanner a 3,000-page allowance. The official site says it checks internal and external URLs and displays where a bad link appears in the HTML. There isn't a separate cap on the number of hyperlinks within those pages.
I'd choose it when a small browser checker stops too early and you don't want a desktop installation. The larger page allowance makes it a reasonable candidate for an older blog with many archived posts. Showing the source location also helps when a link appears in a footer or repeated template.
The catch is that the free edition and the commercial service have different reporting and scheduling features. Don't assume the free scanner includes the CSV exports or recurring scans listed for the commercial version. Ask about the paid service if those are the features you're choosing it for.
A source-code location can also require more work than an editor-level reference. If the link is inside a template, find the template setting or component responsible for it. Editing the generated HTML on one page won't necessarily fix the next page load.
Pick it if a free scan of up to 3,000 pages fits the job. Skip it if you need an ongoing exported report and expect the free edition to provide it.
Ahrefs Webmaster Tools: the owner option for a growing site
Ahrefs gives verified projects 5,000 Site Audit crawl credits per month. Its official free-tools page explains an unusual counting rule: credits are used for HTML pages returning status 200. Broken URLs, redirects and non-HTML resources don't consume those credits.
That makes Ahrefs worth considering when you manage a growing public site and can verify ownership. I'd use the audit alongside the site's other SEO data, then turn each confirmed broken internal link into an editing task. A cloud report can be easier to revisit than a one-off desktop export.
The weakness is ownership and account setup. This free route is for verified websites, so it isn't the easiest way to check a random site's links. It also puts the crawl in an external service rather than on your own computer.
The free allowance renews monthly per project. Broader Ahrefs subscriptions increase access and allowances, but I'd first see whether the owner audit covers the job. Paying for a large SEO suite solely to find a few broken links is hard to justify.
Pick it if you want an audit of a website you can verify and maintain. Skip it if you only need a quick check of a page on someone else's site.
W3C Link Checker: a focused check for links and anchors
The W3C Link Checker looks at links, anchors and referenced objects in a document or stylesheet. Its form also offers recursive checking with a chosen depth. That makes it a different kind of pick from an SEO dashboard.
I'd reach for it when a particular help page has a questionable table of contents or a link that should land on a section. A URL can return a working page while the fragment after the # no longer points where intended. Checking the document's references is a more specific job than collecting an SEO health score.
The form is plain, and the output can involve markup details. I wouldn't hand it to an occasional editor who wants a tidy weekly task list. For someone maintaining HTML, that directness is a benefit: there are fewer unrelated marketing metrics around the question.
It's a free public checker without a paid account requirement. Start with the document you care about and a sensible recursion depth. A narrow question usually produces a report that is easier to interpret.
Pick it if links and in-page anchors are the focus of the check. Skip it if you want a friendly ongoing project dashboard for a whole content team.
SiteOne Crawler: wider reports without a license subscription
SiteOne Crawler is an MIT-licensed project with Windows, Mac and Linux support. Its repository offers an interactive command-line wizard and a separately linked desktop GUI. Reports include HTML, JSON and text, with checks for errors such as 404s and redirects.
I'd consider it when broken links are one part of a wider technical review. You can keep the report local and hand a readable HTML result to someone working on the site. Machine-readable output gives a developer another route for repeatable checks.
The breadth is also its weakness. The project includes auditing, cloning and export features, so a first-time user has more choices to understand than in a dedicated link form. Choose an audit workflow and avoid adding jobs that aren't part of the repair request.
There is no paid license subscription required for the open-source crawler. Hardware, configuration and the website's response still affect what it can scan. I'd treat it as a tool for someone willing to learn the report, rather than the default for an editor making two quick corrections.
Pick it if you want local open-source crawling and wider technical reports. Skip it if a single-page browser checker already answers the question.
For developers: lychee checks the files before they go live
lychee reads links from Markdown, HTML and other input files. Its getting-started documentation makes a key distinction: directories are traversed recursively, but a remote webpage input doesn't become a recursive crawl of the entire website. Supply the pages or built files you want checked.
The surprise is how a working GitHub fragment can fail a checker. In issue 1729, opened June 2025 and still active in 2026, a maintainer explained that some GitHub Markdown headings work through JavaScript without matching HTML IDs. That's a reason to inspect fragment reports before deleting perfectly good references.
I'd choose lychee for a documentation repository or static-site build. You can make the check part of the publishing workflow instead of waiting for a reader to report a dead link. The drawback is configuration: online URLs and local build paths can need explicit mapping. The Screaming Frog vs lychee comparison covers that choice.
Pick it if your content lives in source files and you want repeatable checks. Skip it if you expect one homepage URL to discover and audit every page for you.
Decide what the report needs to contain #
For a content repair, keep source page, failed destination and link text together. The source tells the editor where to work, and the text helps identify the right reference when the page has many links. A bare list of failed URLs pushes that search back onto the editor.
For a static website, decide whether you are checking the current public site or the build you're about to publish. They can contain different files. A successful public URL doesn't prove the next build includes that resource.
Choose one checker that fits the workflow, then make a clear repair list. The step-by-step Screaming Frog guide shows how to turn crawl output into source-page edits.
How we chose #
I checked official feature descriptions, free allowances, documentation and current project activity on October 9, 2026. Recent user discussions and maintainer explanations informed the practical catches, with browser tools favored for occasional owners and source-file checkers kept in the developer section.
FAQ #
What's the best free broken link checker for a small site?
I'd start with Screaming Frog when the site fits its 500-URL free crawl. The source-page report makes it practical to locate and change the links. Dr. Link Check is my browser alternative for occasional scans. Choose by the size of the crawl and the kind of report you'll actually use to make repairs.
Is a free broken-link checker enough for a larger blog?
It can be, if the tool's allowance and discovery method cover the blog. BrokenLinkCheck.com offers 3,000 pages, while Ahrefs provides monthly crawl credits for verified properties. These units aren't interchangeable. Check whether the scan finished and whether older archives were reached before treating the report as a complete list of repairs.
Can these tools find broken external links?
Yes, several picks check external destinations referenced by the pages they scan. Screaming Frog and BrokenLinkCheck.com document internal and external checks. The repair still happens on your own source page: update the reference, replace it with a better destination or remove it when the resource is gone. A report doesn't make that editorial decision for you.
Can a link checker fix the links automatically?
The tools listed here primarily help you find and inspect failures. I'd make the correction in the CMS or source file, then check the changed page again. Automatically replacing a URL with the homepage can hide the error while sending readers somewhere irrelevant. Choose a replacement that answers the same question as the original reference.
Does Google Search Console replace a broken-link crawl?
I'd use Search Console and a link checker for different questions. Search Console provides Google's view of a verified property; a crawler helps trace links in the pages you're auditing. For a repair task, you need the source page containing the bad reference. Keep that source-and-destination information in the task you give an editor or developer.
Sources #
Official tool and allowance links are in the table and individual sections. The exclusive experience sources here are September Screaming Frog feedback and lychee's GitHub fragment discussion.
Featured illustration: WebIndexly, created with OpenAI image generation.














