Find broken links.
Skip the setup.
A tiny browser-based checker for broken internal links, missing files, and image alt issues. Drop one HTML file onto your website and start scanning.
index.html to your website.What it checks
Useful diagnostics without a backend or command-line tools.
Broken internal links
Detect HTTP errors, timeouts and network failures, with the source pages linking to each URL.
Optional file checks
Scan images, documents, scripts, stylesheets, fonts and other linked resources.
Image alt issues
Flag images with missing or empty alt attributes on any checked page.
Parallel crawling
Control concurrency, request delays, timeouts and page limits.
Sortable results
Inspect visited pages, titles, HTTP status codes, size and response time.
HTML report
Pause, resume or stop, then save results as a standalone sortable HTML report.
Install in seconds
Download index.html from GitHub and place it on the website you want to scan.
Open the file in your browser, choose settings and press Start.
Why same-origin?
The browser needs to read page responses to discover more links. Due to cross-origin restrictions, the checker must run on the same origin as the website being checked.
Configurable crawl
Defaults live in CRAWL404 at the top of the HTML file. Main settings are also available in the interface.
| Setting | Default | Purpose |
|---|---|---|
start | / | Start URL |
skip | [] | Do not crawl below selected paths |
delay | 0 ms | Delay per worker |
limit | 1000 | Maximum pages (files excluded) |
parallel | 20 | Concurrent requests |
timeout | 15000 ms | Request timeout |
files | false | Check linked files |
listPages | true | Show visited pages |
listFiles | false | Show visited files |
Good to know
Internal links only
External links aren't checked because browsers can't generally inspect cross-origin responses.
Soft 404s
“Not found” pages returning HTTP 200 count as successful responses.
Response timing
Timing runs until headers arrive; it includes network latency and queuing.
Not a search engine
The checker intentionally ignores robots.txt and sends requests without cookies.
Requires a modern browser. Alpine.js and Bootstrap 5 are loaded via CDN.
One file. Zero setup. Just crawl.
Get the source code, README and MIT license on GitHub.
Open GitHub repository ↗