Guides
Guides
Everything worth knowing about getting things back off the web, whether or not you ever use this tool.
How to download a website
Five methods, from the browser save you already have to the command line, and the one that applies when the site is already gone.
Site downloaders, and how deep to crawl
What a site downloader returns, why depth is the setting people get wrong, and what never comes back no matter the tool.
Wayback Machine alternatives
Six archives and the job each is built for. Includes one the whole internet still recommends that has been offline since 2025.
The Wayback Machine and YouTube
Why the archived page loads but the video will not play, and what the surviving metadata is actually good for.
Finding a deleted YouTube video
What survives a deletion, how to search from a video ID, and how to tell early whether to keep looking.
What the Wayback Machine is
The Wayback Machine is the Internet Archive’s record of public web pages. It takes dated snapshots and keeps them, so a page that changed or disappeared can still be read at an earlier date. It is built for reading those snapshots one page at a time, and it was never built to hand a whole site back to you. That is the part this does.
Coverage is uneven. A popular site may have thousands of snapshots across two decades. A small site may have a handful, and some pages were never captured at all. A restore returns what the Archive actually holds, and no tool can return more than that.
What a web archive can and cannot give back
A web archive stores the files a server sent to a crawler. That means the HTML of each page, the CSS that styled it, the images and scripts the page requested, and the timestamp of the capture. All of that comes back.
What never entered the archive cannot come out of it. There is no database behind an archived page, no server code, no admin login, and no content that sat behind a password. A restored site is the output the archive recorded on the day of the capture.
There are open-source command line tools that do the same job. The Ruby wayback_machine_downloader will download the last version of every file present for a domain into a browsable folder, and waybackpack does the equivalent in Python with date range flags. If you are comfortable in a terminal, use one of those. This exists for the case where you would rather not.
Video is the common surprise. A page that embedded video comes back with the page intact and the video missing, because the file streamed from somewhere the crawler never followed.