SingleFile alternatives

SingleFile is genuinely good, free, open source, and the default recommendation for saving a page as one HTML file. If it works for your pages, use it. People look for alternatives for four specific reasons, and the right alternative depends on which one you have.

Why people look for something else

1. Design-led pages come out wrong. Webfont adoption, pseudo-elements, masks and layered gradients are where fidelity is decided on modern marketing sites, and this is where the gap shows: in our measurements SingleFile scores 99.99% on a Wikipedia article and 99.84% on a simple agency site, but 74.64% on a commercial WordPress theme demo, 81.41% on a page built from SVG filters and 65.55% on stripe.com.
2. They need many pages, not one — SingleFile is a per-page tool.
3. They need the page in a CMS, editable, not as a frozen file.
4. They need it in a pipeline without a browser in the loop.

The alternatives

CopyAnySite — when the page is the design

A Chrome extension that reads the live DOM plus the computed style of every element and pseudo-element, keeps hidden SVG sprite sheets, snapshots canvases as pixels, preserves the document's quirks mode and real body width, and rewrites embedded faces to font-display: swap so a fresh render cannot lose the webfont race. On the same eight-page benchmark it averages 98.87% against SingleFile's 87.93%, and on the three pages above it lands 98.39%, 98.71% and 95.99%.

It also does the things SingleFile does not: bounded whole-site crawls, ZIP bundles, editable WordPress/Gutenberg section blocks, an AI dev kit for coding agents, and an MCP server.

Honest trade-offs: it is not free (one-time licence), and the files are bigger — our stripe.com copy is 50 MB against SingleFile's 5 MB, because embedding everything means a heavy page stays heavy. SingleFile also left zero remote requests across the benchmark against our four.

Choose it if: the copy has to look right offline on a design-led page, or has to land in WordPress.

monolith — a CLI, no browser

Rust, one command, one file. Great inside scripts and CI. It fetches the server's HTML, so client-rendered content is missing, and in our runs it left 365 requests aimed at the original hosts across eight pages, averaging 83.44%.

Choose it if: you need scripted bundling of server-rendered pages.

wget -p -k — already installed

Free, everywhere, and fine for a simple page or a whole traditional site. Produces a folder rather than a file, needs the right flags to convert links (87.40% average, 554 remote requests left in our runs), and cannot see JavaScript-built content.

Choose it if: you are in a terminal and the site is server-rendered.

HTTrack — whole sites, with a GUI

The classic mirroring tool, with rules and filters and a Windows front-end. Same server-fetch model and the same blind spot on client-rendered sites.

Choose it if: you want hundreds of pages of an old-school site without writing flags.

Chrome's built-in save

Ctrl+S "Webpage, Complete" gives you an HTML file plus a folder; MHTML (behind a chrome://flags switch) gives you one file that mainly Chrome opens. Both save less of the rendered state than any extension listed here.

Choose it if: you need something in five seconds with nothing installed.

Print to PDF

Not a page any more, but flat, portable and unarguable — the right answer when the goal is a record rather than a working page. Check what @media print removed.

How we measured

Eight pages, three runs, median. Each copy is rendered with the original host unreachable and diffed against the live page with pixelmatch at desktop and mobile widths; requests still pointing at the origin are counted. The harness lives in our repo and runs with npm run bench:rivals. We publish the rows we lose as well as the ones we win — on Hacker News, monolith and wget also reach 100%.

Quick answer

• Article or docs page, free → SingleFile.
• Marketing page, theme demo, anything with real visual craft → CopyAnySite.
• Whole traditional site → HTTrack or wget --mirror.
• Scripted, no browser → monolith.
• Just a record → PDF plus a Wayback submission.

FAQ

Is there a better alternative to SingleFile?

For text-heavy pages, not really — it is free and near-perfect there. For design-led pages, tools that capture computed styles and pseudo-elements score substantially higher; our benchmark puts the gap at 13–30 points on those pages.

Is SingleFile free?

Yes, free and open source, for Chrome and Firefox, with no page limits.

What is the main limitation of SingleFile?

It serialises the DOM and inlines assets, but does not fully reproduce the computed visual state — so webfont adoption, pseudo-element content, masks and complex gradient stacks are where its copies drift from the original.

Can SingleFile save a whole website?

No, it is a single-page tool. Use a crawler for multiple pages.

Does SingleFile leave links pointing at the original site?

Very few — it inlines assets at any size, which is why it left zero remote requests in our benchmark. That is a real advantage over the CLI mirroring tools, and over our own captures' handful of oversized-media exceptions.

Keep reading

HTTrack alternatives

HTTrack still mirrors traditional sites well and cannot see JavaScript-built pages at all. The alternatives, by what you are actually mirroring.

wget alternatives for downloading a website

The flags that make wget work, the sites where it cannot work at all, and what to use instead for each case.