// How it works
Paste a page address to search for that page, or a site address to search more broadly.
Crawlback checks Common Crawl's index for saved versions, with their capture dates. An empty result does not prove a page never existed.
Open matching snapshots, check them against other sources, and save the results if you want.
// Why Crawlback
Finds snapshots the Wayback Machine can miss, including moments it never captured.
Searches run from your browser, and your history stays there with no server-side storage. You can wipe everything from Settings.
Archived HTML renders in a sandbox. No archived scripts execute, nothing phones home.
Export capture lists as CSV or download the raw HTML as a ZIP file to keep.
// Data source
Common Crawl is an independent nonprofit that has been crawling the web since 2008 and publishes everything it collects as open data. Its archive spans more than 300 billion pages across 100+ crawl collections. It collects samples of the web, so it does not save every page or capture every change.
Crawlback uses those crawls to find individual pages inside the archive. Results are captured HTML, not live replays of the page.
// FAQ
The Wayback Machine archives on demand and replays full pages. Crawlback searches Common Crawl, a separate archive built by broad crawls since 2008. The two overlap only partially. Common Crawl often holds captures the Wayback Machine is missing, which is exactly when Crawlback is useful.
Common Crawl limits request rates, and everyone on the same network can share one public IP address and its limit. Crawlback slows down on purpose to avoid getting your search blocked, so some scans take 20 minutes or more. If you get an HTTP 503 response, slow down. If Common Crawl temporarily blocks your IP address, wait 24 hours before trying again. Avoid parallel requests. Common Crawl's public index API is often overloaded, so if you get many errors, simply try again later.
The preview is incomplete for security and privacy reasons. Common Crawl mostly stores the captured HTML, not the images, stylesheets, and scripts that were loaded with it at that moment. Crawlback shows the captured HTML in a sandbox: archived scripts never run, so a capture cannot execute code. The preview also does not load external resources from the original site or third-party servers, so opening a capture does not contact those servers or reveal your visit or IP address to them.
Keep the exact URL and the capture date. The capture date is when Common Crawl saved the page, not when it was first published. Compare the capture with other captures and independent sources. A capture alone does not prove who wrote a page or that its claims are true.
Crawlback checks whether Common Crawl is currently blocked or opted out for that URL. Even if it is blocked now, you can still find older snapshots from times when the page was crawlable.
Almost never. Your browser sends archive searches straight to Common Crawl's public index and data servers, not to the site you're researching. Crawlback's server makes one request to the target, a robots.txt check. For sensitive research targets, you can turn that off in Settings so the website is not contacted at all.
Crawlback is open source. Read the code on GitHub →