Turning a PDF into a web page without getting a screenshot

The short answer

To turn a PDF into a web page, convert it to HTML with a tool that extracts the text and writes it as real HTML elements, not one that renders each page to an image and wraps it in an img tag. Then check the result: if you can select a single word and search for it in the browser, you have a web page.

That check matters because both approaches produce a file ending in .html that opens in a browser and looks like the original. Only one of them gives you something a search engine can read, a screen reader can speak, or a phone can reflow.

Two things that both get called PDF to HTML

A PDF stores text as positioned glyphs: a font, a string of character codes, and coordinates on the page. Getting that into HTML means reading those runs out, working out which ones belong to the same line or paragraph, and writing them back as elements the browser lays out itself. It is real work, and the result is never pixel-perfect, because the browser is now in charge of rendering the type.

The shortcut is to render each page to a bitmap, exactly as a PDF viewer would, and emit an HTML file containing those images. It looks flawless, because it is a picture of the page. It is also not text. Nothing can be selected, the browser's find command finds nothing, a search engine sees an empty page with some images on it, and the file is usually far larger than the PDF it came from.

Some converters sit in between: an image of the page for appearance, with an invisible text layer positioned over it for selection. That is better than a bare image, but it inherits most of the image approach's weight and none of its responsiveness.

How to tell which one you were given

Open the HTML file in a browser and run through these. Each takes seconds.

• Select a word. Real text highlights word by word. An image either selects as one block or not at all.
• Use the browser's find command on a word you can see. If it reports no matches, the text is not in the page.
• Zoom to 300%. Real text stays sharp at any zoom. A rendered page goes soft and blocky, because its resolution was fixed when it was rendered.
• Look at the file size. A text-based HTML file is typically smaller than the PDF. One that is several times larger is almost certainly carrying page images.
• View the source. Paragraphs of readable words mean text. A wall of img tags or long base64 strings means pictures.

Layout fidelity versus a page that reflows

Even among converters that emit real text, there is a choice to make. A PDF page has a fixed size and every line has a fixed position. A web page has neither.

One option keeps the positions: each line of text becomes an element absolutely placed at the coordinates it had on the PDF page. The result looks close to the original and the text is selectable, but it behaves like a printed page on a screen. On a phone you pinch and pan; nothing wraps.

The other option discards positions and emits ordinary structure: headings and paragraphs, in reading order, that the browser wraps to whatever width it has. It will not look like the PDF, but it reads properly on a phone, works with screen readers, and is the version worth publishing. Headings have to be inferred, usually from font size, because most PDFs do not record which lines were headings.

World of PDF's PDF to HTML tool offers both: a layout mode that positions each line where it sat on the page, and a reflow mode that writes headings and paragraphs. Both emit real text, not images, and the conversion runs in your browser rather than on a server. Neither carries the PDF's images across, so a document whose meaning lives in its figures needs more than a conversion.

Scans have no text to convert

If the PDF is a scan or a phone photo, the page is already an image. There are no characters in the file, only pixels shaped like them, so a text-based converter has nothing to extract and an image-based converter just hands the picture back.

The fix is OCR first: recognise the characters and add a text layer, then convert. Expect to proofread the result, because OCR is a best guess and errors that are easy to ignore in a PDF become obvious in a web page.

When to just link the PDF instead

Converting is not always the right move. If the document is a form, a contract, a technical drawing, a brochure where the design is the point, or anything people will print, the PDF is already the correct format. Upload it, link to it with a descriptive link text, and put a short HTML summary on the page so search engines and readers know what it contains.

Convert when the content needs to live as part of the site: an article, a guide, documentation that will be updated, or anything that most people will read on a phone. In those cases, use the reflow output as a first draft rather than a finished page. Fix the headings, restore the images and links the document needs, and the result will outperform the PDF on every measure that matters for a web page.

Frequently asked questions