Add header & footer

Rows are placed 28 points in from the visible edge of the page: the first header baseline that far below the top edge, the last footer baseline that far above the bottom edge, with additional lines stacking inwards at 1.25 times the font size. That margin is fixed and there is no control to widen or narrow it. The live preview is drawn in the same points the export uses rather than in screen pixels, so the gap you see on screen is the gap you get on paper. If a printer with a wide unprintable border clips a row, the practical fix is to add a blank first line to the header — or a blank last line to the footer — which pushes the whole block further inwards. Mixed page sizes are measured individually.

Mirrored odd-and-even rows are not available: whatever you type goes on every page, with a single alignment for the header and a single alignment for the footer, and there is no option to exclude the first page. For a facing-page layout where the running head flips to the outer edge, the workable route is to split the document into contiguous parts with Split PDF, stamp each part with its own alignment, and reassemble with Merge PDF — which only helps when the sections you want treated differently are consecutive ranges rather than alternating pages. The same trick covers a cover page that should stay clean: stamp the body on its own and merge the cover back afterwards. For double-sided printing, a centred row is the safest single choice.

The four placeholders are resolved the moment you press the button. {n} is the page's own number and {total} is the document's page count, so Page {n} of {total} counts every page with no offset available. {date} is today's date taken from your computer, formatted as YYYY-MM-DD — it records when you stamped the file, not when the PDF was created, and it does not refresh later. {filename} is the name of the PDF you dropped in with the .pdf extension removed, which is what makes printed handouts traceable back to a source file. Everything outside the braces is printed literally, so they mix freely: {filename} — {date} — page {n} of {total} is a valid single row.

Accented and non-Latin text is handled for you. Before drawing anything, the tool resolves the exact rows it is about to stamp and inspects them: if every character fits the standard Latin encoding it uses built-in Helvetica, and if anything falls outside it — a Polish ł, Cyrillic, Greek, Hebrew or Arabic — it switches to an embedded Unicode font so the run does not fail after you have finished configuring. Only text that is actually drawn gets inspected, which means a non-Latin filename does not pull in the larger font unless a row genuinely references {filename}. Embedding that font adds its glyph data to the saved file. If a character still cannot be drawn, the error line under the card says so rather than silently dropping it.

Nothing is uploaded when you stamp a header or footer with this tool. The PDF is read from your disk by the browser, the page count and every resolved row are worked out in the tab you have open, and the stamped copy goes straight to your downloads under a -headerfooter name. No copy is kept anywhere, because there is nowhere for one to be kept — this tool involves no account, no storage and no processing server. That matters for the sort of document that usually needs a running head: internal policies, contracts under negotiation, HR letters, board packs. It also means the text you type stays in your browser, which is worth knowing when the header itself is the confidentiality classification you are trying to protect.

Add header & footer is free and unmetered: no page cap, no daily quota, no signup, and no premium tier that unlocks extra alignments or placeholders. Every page is stamped in the same run, whether the file has two pages or two hundred — long documents just take a little longer, because each page is measured and drawn individually. The output carries only the rows you typed. PDFChamp does not append a footer of its own, does not write a credit line into the document metadata and does not watermark the result, so what you download can go to a client exactly as it is. Either box may be left empty when you want a header alone or a footer alone.