PDF Converter

This is the most important thing to know about converting out of a PDF, and it is a deliberate design choice rather than a limitation to apologise for.

What survives: the text, in reading order, with headings recognised by their size, paragraphs kept intact, tables reconstructed into rows and columns, and lists preserved. Open the result and you have a document you can edit, search and reuse.

What does not: the original page design. Exact fonts, precise spacing, multi-column layouts, floating images, headers and footers, and pixel-level positioning are not reproduced. A visually elaborate brochure converted to Word gives you its words in a sensible order, not a Word file that looks like the brochure.

The alternative would be a document made of text boxes positioned absolutely on the page — which looks closer at first glance and is miserable to edit, because changing one sentence breaks the layout. Producing something editable is the more useful answer, and the honest one.

A scanned document is a photograph of a page. It contains no text at all — no glyphs, no fonts, nothing to read out — so converting it to Word, Excel or plain text correctly produces an empty document.

That is not a failure of the conversion; there was genuinely nothing there. The quick test is to open the PDF and try to select a sentence. If nothing highlights, it is a scan.

Recovering text from a scan needs optical character recognition, which reads the picture and guesses at the characters — a different operation with its own accuracy trade-offs, particularly on handwriting, poor scans and complex layouts. What does work on a scan is PDF to Images, which gives you the pages as pictures, and Extract PDF images, which gives you the original scans at full resolution.

This direction is generally more faithful, because it goes from a structured document to a fixed layout rather than the other way round. Word documents, presentations, spreadsheets, OpenDocument files, RTF, Markdown, plain text and EPUB all convert in, along with JPG, PNG, WEBP, SVG and HEIC images.

Fonts are the usual source of difference. A document set in a typeface the conversion doesn't have will be rendered in a substitute, which changes line breaks and can shift pagination. Sticking to common fonts, or converting from the application that created the file where fidelity is critical, avoids it.

Some formats are deliberately not offered, for the same reason the rest of the site avoids overpromising: the legacy binary Office formats (.doc, .ppt, .xls) need a full office suite to read properly, and TIFF has no browser decoder. A tile that produced a poor result would be worse than no tile.

Both directions are computed locally. PDFs are parsed with pdf.js and written with pdf-lib; Office and OpenDocument files are ZIP archives of XML, which are unpacked and rebuilt in the tab. Nothing is uploaded, so there is no file size limit, no queue, and no copy of your document sitting on a server afterwards.

The practical ceiling is memory. A very large document converted to a very large output holds both in the tab at once, so a phone gives up sooner than a laptop does. Converting in sections is the workaround, and Split PDF is how to make the sections.

The code for a given conversion is only downloaded when you ask for it, so opening this page does not load the machinery for all thirty-odd formats.