How alternate-mix works

1. Upload your odd-pages PDF. This is the file your scanner produced on the first pass — pages 1, 3, 5, 7 and so on. It loads into the browser, pdf-lib counts the pages, and the file goes nowhere else.
2. Upload your even-pages PDF. This is the file from the second pass — pages 2, 4, 6, 8. The order depends on how you flipped the stack: most automatic document feeders produce these in reverse (last even page first), but some keep natural order.
3. Pick the order mode. Forward (A1, B1, A2, B2 …) when the even pages are in natural order. Reverse (A1, B(last), A2, B(last-1) …) when the even pages came out backwards. If unsure, run it both ways and check page two of the output — it should be the back of page one of the original.
4. Click Combine. pdf-lib walks both page lists and writes the interleaved sequence to a new in-memory PDF. The output download link appears below the button.
5. Download and verify. Open the file. Pages should now read 1, 2, 3, 4, 5, 6 — the original two-sided document reassembled. If the page count looks wrong (one stranded page at the end), the inputs had unequal counts; the warning above tells you which file was longer.

Why a separate tool, not just /merge?

The plain /merge tool concatenates: file A's pages then file B's pages. That gives you 1, 3, 5, 7, 2, 4, 6, 8 — useless for reading. To get 1, 2, 3, 4, 5, 6, 7, 8 you have to interleave: take one page from A, then one from B, then back to A, then back to B. Doing this manually in /merge would mean splitting each file into single-page PDFs first, which is a 5-minute job for a short document and a 30-minute job for a long contract. /alternate-mix-pdf does the interleave in one button press because the only configuration that varies is forward-versus-reverse for the even-pages file.

Sejda is the only mainstream tool that ships an alternate-mix mode under that name. iLovePDF, Smallpdf, and PDF24 all expect you to do the interleave manually inside their drag-and-drop UI. We took Sejda's pattern, kept the forward-and-reverse choice that matters for real scanner output, and skipped the upload step that defeats the privacy purpose for the document types this is most useful for.

Who this is for

Anyone with a sheet-feed simplex scanner trying to digitize two-sided documents. That's a lot of people: legal teams scanning contracts, accountants scanning bank statements, HR teams scanning signed offer letters, students scanning library books, archivists processing back-issue magazines. The pattern is always the same — feed the stack once, flip it, feed it again, and end up with two PDFs that need to be woven back together into one readable document.

India angle: a lot of small offices in India still run older multifunction printers with simplex-only ADFs because duplex models cost noticeably more. This tool is the missing piece between those cheaper scanners and a clean digital archive of two-sided originals — bank statements, GST invoices, court orders, property documents.

Edge cases the tool handles honestly

Mismatched page counts. If one PDF has 39 pages and the other has 40 (very common when the original document had an odd page total), the tool warns you and still produces a 79-page interleaved output with the longer PDF's last page sitting alone at the end. You can inspect and decide whether the stranded page is correct.

Empty file. If you accidentally upload a 0-page PDF (rare but happens when a scanner crashes mid-job), the tool refuses to combine and shows an error pointing at which slot was empty.

Encrypted PDFs. The tool loads PDFs in ignore-encryption mode for the page-count read, which works for most scanner output. If your source has a real password, run /unlock-pdf first, then come back here.

Form fields and bookmarks. pdf-lib's copyPages preserves visible page content but does not carry document-level outlines or AcroForm catalogs across. For scanner output this never matters; for source PDFs that already had bookmarks it does, and you'd want to rebuild the outline afterwards.

Large files. Hundreds of pages on each side is fine. Thousands of pages may exhaust browser memory; split with /split first, alternate-mix the halves, then /merge the results.

Privacy callout

Both files are read by your browser, processed in memory by pdf-lib, and offered back to you as a download. Nothing is uploaded. There is no server-side processing in this tool — open the Network tab in DevTools while you run a Combine and you'll see no outgoing PDF data. This matters more than usual for alternate-mix because the inputs are typically two-sided scanned documents: bank statements, signed contracts, lease agreements, scanned identity documents. Uploading those to a free online tool is a security and privacy hazard. Doing the interleave in your browser keeps every page on your device.

After you've combined

Two follow-up tools are usually relevant. /remove-pdf-metadata strips author, dates, and producer fields from the combined file — useful before sharing a contract or a confidential document. /compress shrinks the file size without re-OCR'ing or re-scanning, helpful when the interleaved output is large because both inputs were image-heavy. If you're handling a scanned document and want to make the text searchable, /ocr-pdf will run Tesseract over the pages in your browser.