Tiny Online Tools
What an MHT file is
When Internet Explorer, Word or Chrome's Save as MHTML wrote a page to a single file, it produced an RFC 2557 email message. One multipart/related body holds the HTML plus every image, stylesheet and script the page referenced, each part tagged with the URL it came from. It is a perfectly reasonable format that almost nothing opens any more.
How it is rendered
The parts are separated, each one is decoded from its transfer encoding (usually quoted-printable for text, base64 for images), and then the root HTML's references are rewritten. That last step is the whole job: every src, href and CSS url() still points at the original absolute address, so without rewriting them the page renders bare. Each one is matched to the part that was saved for it — by Content-Location, or by Content-ID for cid: references — and pointed at that part's bytes instead.
Nothing is fetched, nothing runs
The rebuilt page goes into an iframe with sandbox="allow-same-origin" and nothing else. There is no allow-scripts, so no script in the archive can run.
A Content-Security-Policy of default-src 'none' is injected into the page as well. This matters more than it sounds: any reference the archive did not save stays an absolute http URL after rewriting, and a browser rendering it would quietly fetch it. The policy stops that. Only blob: and data: URLs — the parts that are inside the file — are allowed to load.
Incomplete archives
Web archives are saved incompletely all the time: a background image declared in a stylesheet the saver did not follow, a lazily loaded picture, a font on another domain. Those references are counted and listed. If the page looks wrong, that list usually explains why, and it is a far better answer than a broken image icon.
Also shown
The subject, the date the page was saved, and the original URL — the root part's Content-Location, which is where the page actually came from. Plus every part with its MIME type, transfer encoding, declared charset and decoded size, and a toggle to read the root HTML source.
Privacy
The archive is opened in your browser. It is not uploaded, and neither it nor the tool contacts the network.