PDF tools that never see your file

What actually survives an edit

A PDF is not one thing. Alongside the visible page content it carries an outline (bookmarks), form fields and their values, tag structure for screen readers, file attachments, digital signatures and document metadata. Most online tools quietly discard several of those, and you only find out later.

Here is what ours do, and the dividing line is sharper than you might expect. Merging, splitting and page extraction build a brand-new document and copy pages into it. Everything that lives *on a page* comes across exactly — text, images and vector artwork are copied, never recompressed or rescaled, so the pages are visually identical. But everything that lives at the *document* level stays behind: the bookmark outline, the form dictionary, the tag structure a screen reader follows, the document language, and the Title and Author fields. A merged or split file therefore has no bookmarks at all — not the first file's, not any — and a form's boxes may still draw while no longer being fillable fields. If preserving an outline matters, qpdf keeps the primary document's catalogue when it merges; nothing that runs in a browser does.

There is no resource sharing between inputs either, which explains a result people find surprising: merge two reports that embed the same font and the font is embedded twice, so the combined file is roughly the sum of its parts rather than smaller. Rotation is the exception to all of this — it loads the document and re-saves it rather than building a new one, changing only the page's rotation entry, so it is lossless, reversible, and keeps the outline, form, tags and metadata intact.

Compression is the one that trades something away, and it is honest about which: images are re-encoded at a lower quality, while text and vector artwork are left alone. That is why compressing a text-only report saves almost nothing and compressing a scanned brochure can save ninety per cent.

A digital signature does not survive any edit at all — by design. Signing covers the exact bytes of the document, so merging, splitting, rotating or compressing a signed PDF invalidates the signature. That is the format working correctly, not a bug in the tool, and no PDF editor anywhere can avoid it.

Where a browser is genuinely the wrong tool

Running in the browser is what makes these private, and it costs something. We would rather say so here than have you discover it halfway through a job.

Very large files. Everything is held in your device's memory, so the ceiling is the tab's memory budget, not a plan limit. A few hundred megabytes is comfortable on a laptop; a phone will give up sooner. A server-side tool can stream a two-gigabyte file that we cannot open.

Scanned pages. If a PDF is photographs of paper, there is no text in it to find. Our OCR tools read those pages well enough to search and copy from, but a dedicated desktop OCR package trained on your language and layout will beat them, and neither will reconstruct a complex table perfectly.

Exact print reproduction. Spot colours, overprint, trapping and ICC rendering intents are the domain of professional prepress software. We preserve the colour data already in your file, but we do not manage it.

Batch and automation. These tools run when a person clicks. Converting four thousand invoices on a schedule is a job for a command-line tool on a server — Ghostscript, qpdf and pdfcpu are all free, and better at it than any web page.

Choosing between the tools that sound alike

Merge PDF vs Rearrange PDF pages. Merge joins whole files and lets you order the files. Rearrange works inside one document and lets you drag individual pages. Combining several files *and* interleaving their pages is two steps: merge, then rearrange.

Split PDF vs Extract pages. Split cuts one document into several documents. Extract pulls a selection out into a single new file and leaves the original alone.

Unlock PDF vs Remove restrictions. Unlock needs the password and removes the encryption. Restrictions are the owner-set flags that disable printing or copying; they are advisory and can be cleared without a password. Neither is a way into a document you are not entitled to open.

PDF to Word vs Extract text. PDF to Word rebuilds paragraphs, headings and tables as editable Word content and will be imperfect on complicated layouts. Extract text gives you the raw characters with no formatting, which is what you want for search, analysis or feeding something else.