Alternating two PDFs means combining them so page 1 of document A is followed by page 1 of document B, then page 2 of A, then page 2 of B, and so on; once one document runs out first, every remaining page from the longer document is appended in its original order, and the entire operation stays inside your browser tab. The rule is fixed, visible, and matches a narrow kind of real work. Use it only when your reading order is exactly A1, B1, A2, B2 with unmatched tail pages preserved, and both input documents are already arranged the way each one should be read on its own. It is not simple concatenation (every page of A followed by every page of B), it is not reversal, and it is not range picking; for those jobs a merge tool or a reorder tool is the better fit. Each input must be a non-empty PDF no larger than 25 MiB, and the output filename follows a [a-name]-[b-name]-alternated.pdf form so you can tell at a glance that the document was interleaved rather than merged. Knowing this exact contract up front is what lets you answer the question of whether the tool matches your task.
What "alternating PDF pages" actually means
The interleaving rule is intentionally simple and documented up front, so you can decide whether it matches your task before clicking anything. After you pick the first file (always labeled A) and the second file (always labeled B), the tool copies pages in this sequence: A1, B1, A2, B2, continuing while both files still have a page at the current index. The instant one of the two files runs out first, the remaining pages from the longer file are appended in their original reading order, not skipped, reversed, or moved up. Two concrete examples from the tool's documented behavior: a three-page A combined with a two-page B produces A1, B1, A2, B2, A3; a one-page A combined with a three-page B produces A1, B1, B2, B3. Both examples illustrate the same rule, which is that pairing happens index by index and any leftover pages from the longer document tag along at the end untouched.
Pages are copied as real PDF page objects through the underlying copyPages operation rather than as flattened screenshots, so the copied page keeps its own page-box dimensions, rotation flag, vector drawing, text content, and embedded raster images to the extent the library supports copying. The tool does not force every output page to the first document's size, does not stretch content, and does not insert any blank separator between paired pages. The result line in the interface reports the page count of each input alongside the total output page count, and that total must equal the sum of the two input page counts by construction; if a result line ever shows a mismatch, the workflow has not finished and the output should not be treated as final.
Five situations where alternating is the right choice
When the rule above matches your reading order, the job is a fit. Five common reader situations line up well, and each one is the kind of task where the desired truth is the documented A/B alternating sequence.
- Duplex scans split into two stacks. A reversing scanner sometimes saves pages 1, 2, 3, … in one PDF and pages …, 3, 2, 1 reversed in another, or splits front and back sides into separate files. If the two stacks are already each in their intended internal order, alternating produces one continuous spread-ready PDF.
- Question and answer documents. Many teachers and exam authors keep the question set in one file and the answer set in another so they can print the blank version separately. Alternating them gives a single key-sheet PDF for review, while keeping the original question-only and answer-only files untouched.
- Translated pages alongside originals. When a translator delivers a mirrored bilingual draft (page 1 of source, page 1 of translation, page 2 of source, page 2 of translation), alternating is exactly the production step that builds the combined handout.
- Two chronological page sets. If event A and event B each have their own date-stamped PDF and the merged story is "A page one, B page one, A page two, B page two", alternating pairs them index by index with any later extras placed at the end.
- Odd and even batches already in order. Scanners that batch odd-numbered pages into one file and even-numbered pages into another, with each batch already internally ordered, are a textbook fit. The tool will not reorder, deduplicate, or guess which batch is "front" versus "back"; as long as the batches are correct, the output is correct.
Before you click, walk through one quick self-check: is the desired reading order really A1, B1, A2, B2 with any leftovers appended? Will every page from both inputs appear in the output, just in the new order? Do both inputs look correct to you on a casual page-by-page glance? If yes to all three, the interleaving tool fits the task. If the order is different, or you want some pages omitted, the tool is the wrong one and a different PDF tool will fit better.
When you should reach for a different PDF tool
Several neighboring tasks look similar but need a different operation. If the desired output is every page of A followed by every page of B, you are describing a concatenation job, not an interleaving job. If you want a fully custom new sequence, moving individual pages around, dropping some, or repeating others, you want a sequencing tool instead. If your end product is a folded printed booklet where every two source pages must be placed on the same physical sheet in imposed order, use the booklet tool, because alternating does not create signatures, and booklet imposition does not interleave index by index.
If one of your two inputs came out of the scanner reversed, fix that input first with a reorder or reverse tool and only then mix the corrected files; the interleaving tool is explicit that it keeps existing internal order and does not reverse either source.
| If your desired output is | Reach for |
|---|---|
| Every page of A followed by every page of B | Merge PDF |
| A fully custom page sequence across files | Reorder PDF Pages |
| Folded-booklet printing signatures | Booklet PDF |
| A1, B1, A2, B2 with any unmatched tail preserved | Alternate Mix PDF |
| Selected pages copied out, in a chosen order | Extract PDF Pages |
| A page mirrored or flipped | Flip Mirror PDF Pages |
In short: do not use the alternating tool for concatenation, range selection, deduplication, content comparison, repair, booklet imposition, or reversal. Use it only for the documented A-then-B index pairing with unmatched tail pages preserved.
How to interleave two PDFs with Alternate Mix PDF
- Open Alternate Mix PDF in your browser tab. The whole job runs locally, so there is nothing to sign into and nothing gets uploaded.
- Pick the first source PDF from your device as PDF A, then pick the second as PDF B. The first file you select is always A and supplies the first page of the output. Confirm both files are non-empty PDFs no larger than 25 MiB; encrypted, damaged, oversized, or mislabeled files surface a visible error and do not proceed.
- Open both source files in another viewer if you have any doubt about internal order. The tool keeps the existing order of both inputs, so fix any reverse-scanned batch with a reorder tool before you mix.
- Click the "Alternate and mix PDFs" action. The page-count result line appears, showing each input's page count and the total output page count, which must equal the sum of both inputs.
- If the page-count total matches the sum, save the new PDF. The filename follows a [a-name]-[b-name]-alternated.pdf pattern that flags the file as interleaved rather than merged, and the original files remain unchanged.
A few operational notes matter when this tool is doing the actual work in your tab: choosing a replacement for either input clears the old output, a job identifier prevents a slower earlier operation from overwriting a newer one, and swapping which physical file you select as A versus B turns the result into B1, A1, B2, A2 under the original slot labels. Memory cost can exceed the sum of both source file sizes during the job, because both documents and the newly serialized result may coexist temporarily. Close other heavy tabs if your device is constrained and keep irreplaceable originals somewhere outside the browser workflow. The 25 MiB limit applies separately to each input, not to their combined output.
Confirming the interleaved result matches your plan
After the result downloads, spend a minute validating that the document is actually what you needed, because the interleaving rule is rigid and silent about content mistakes. A practical pattern for that confirmation is laid out in how to avoid mistakes with alternate PDF pages; the mechanical checks below are a tight version.
The mechanical checks you can do quickly:
- Output page count equals the sum of input page counts. If it does not, the job did not finish or a file was misread; treat the file as not final.
- Page 1 of the output really is page 1 of A. If it is page 1 of B, the two input slots were mentally reversed and a re-run with the files renamed or swapped will produce the intended sequence under a different filename pair.
- Open the file mid-way and confirm the pattern A2, B2, A3, B3 holds wherever both inputs still have pages.
- At the tail, confirm every remaining page from the longer document appears in its original order, with no insertions, blanks, or drops.
Content checks need a little more care.
- Interactive elements such as form fields, bookmarks, outline destinations, scripts, attachments, and cross-document links may not survive page copying with the same semantics, so inspect any interactive PDF after mixing rather than assuming the behavior carried over.
- Each copied page should still have its original dimensions and rotation, because the library does not stretch or resize copied pages; if a page looks the wrong size, the source was that size, not the output.
- If you ever need a different page count, different order, different rotation, or different page sizes, that is a different workflow; alternating alone will not deliver it.
If the result looks right, store the new PDF alongside the two originals; the originals are unchanged. If the result is not what you needed, you have not lost the inputs, and you can re-run with one input reordered, one input replaced, or a different pair entirely without losing work.