Is my file uploaded anywhere?

No. The page reads the file with the browser's own File API and has no network code of any kind. You can verify that in the network panel: after the document loads, nothing else is requested, before or after you drop a file on it.

What counts as self-contained?

Zero external references: no script, stylesheet, font, image, iframe or CSS url() pointing at another host. Everything the page needs arrived with the document itself.

Can I keep it or host it myself?

Yes. Take the file with the download button and put it anywhere you like — it is a single self-contained page, so a copy is a deployment.

Goldenboy · Specimen · free

Specimen 01 · the self-containment check

A small tool, given away, that does one thing: tell you whether an HTML file really carries everything it needs. It runs in your browser, uploads nothing, and is itself a single file that makes no network requests.

01The check

Drop any HTML file below. It reports the file's size, its gzipped size, every external reference it still carries, and how much of it is embedded data. Nothing is uploaded — this page has no network code at all, and the file never leaves your machine.

Drop an HTML file here or choose one — nothing is uploaded

02What it is looking for

Remote subresources the browser would fetch: script src, images, iframe, video, audio, source, SVG use, and link tags whose rel actually loads something — stylesheet, icon, preload, manifest.
CSS url() references to remote hosts, in <style> blocks and in inline style attributes.
@import rules pulling in a remote stylesheet.
Inline script and style blocks, and embedded data URIs — what a self-contained file has INSTEAD of those references.

It parses the document rather than grepping it, and it counts ordinary hyperlinks separately from subresources. A link to another site is navigation and costs a visitor nothing; a remote script is a runtime dependency on somebody else's server. Reporting them as the same thing is the mistake this page exists to correct — zero fetched references is the definition of self-contained, and a page may link anywhere it likes.

Nothing in the file you drop is loaded, executed or requested. It is parsed into an inert document with no browsing context, examined, and discarded.

03This page is its own example

Open devtools, switch to the network panel, and reload. You will see the document and nothing else: no stylesheet, no font, no script, no analytics tag — this is the one page on the site that carries no analytics, because the claim is only worth making if it is true. Save it with Ctrl+S and it still works from your desktop, offline.

Take the file See the objects

04Try it on ours

The objects sold here measure 704,811 bytes (Object 01) and 548,530 bytes (Object 02) on disk, with zero fetched references each. If you buy one, run it through this page: the byte count should match exactly. The gzipped figure will land near our published 187,269 and 403,717 bytes without matching them, because your browser compresses at its own default level and ours were measured with gzip -9. Then run your current site through it and see what it is actually fetching.

What the format is and why it is worth the tradeoffs: single-file websites. What ships without a toolchain: no build step.

FAQQuestions

Is my file uploaded anywhere?: No. The page reads the file with the browser's own File API and has no network code of any kind. You can verify that in the network panel: after the document loads, nothing else is requested, before or after you drop a file on it.
What counts as self-contained?: Zero external references: no script, stylesheet, font, image, iframe or CSS url() pointing at another host. Everything the page needs arrived with the document itself.
Can I keep it or host it myself?: Yes. Take the file with the download button and put it anywhere you like — it is a single self-contained page, so a copy is a deployment.