Combine several PDFs into one document, in the order you choose, without sending them anywhere. Pages are copied rather than re-rendered, so nothing is recompressed, and nothing is made smaller either.
Runs in your browser. Nothing is uploaded.
A PDF is a collection of numbered objects, page trees, fonts, images, content streams, tied together by a cross-reference table that says where each one lives in the file. Merging does not open your documents and paste one into the other. It creates an empty document, copies the page objects across from each source, renumbers everything so the references still point at the right places, and writes a fresh cross-reference table. The bytes that describe your pages are moved, not reinterpreted.
That is why the result comes back at almost exactly the size it went in. Combining a 12-page scan with a 12-page text document measured 10,374,682 bytes in and 10,374,095 out. A difference of 587 bytes, which is the old cross-reference tables being replaced by one new one. It is also why the whole thing takes 32 milliseconds: no page is ever rendered, no image is ever decoded, and no text is ever re-typeset. If you were hoping the merge would also shrink the file, it will not, and any tool that claims otherwise is quietly re-encoding something.
Choose your PDFs
The same questions a security review would ask, answered plainly.
What a local merge gives you that a server-side one cannot.
The file list is the page order. Drag a document up or down before merging and the output follows, so you do not have to rename files into alphabetical submission first.
Because pages are copied rather than rasterised, the merged document keeps selectable text, working search, and any embedded fonts at their original quality.
Add as many PDFs as you like in one pass, and drop individual ones from the list without starting over. There is no per-file or per-merge ceiling to buy your way past.
A 24-page merge of a scan and a text report completed in 32 milliseconds. Structural copying is cheap; the slow part of most online mergers is the round trip, not the work.
There is no upload step and no server to receive one. The page loads once, then all work on your PDFs happens in local JavaScript.
No daily limits, no per-file ceilings imposed by a server tier, no watermarks, and no account. The practical limit is your own device's memory.
Real documents, real byte counts, and what changed between them.
The common case: a signed scan that needs a covering document in front of it.
Two files
12-page scan (10.3 MB) + 12-page text PDF (50 KB)
One file
24 pages, 10,374,095 bytes, 32 ms
The output is 587 bytes smaller than the two inputs added together. Nothing was compressed. That difference is two cross-reference tables being replaced by one. The scanned pages are still the same images at the same resolution, and the typed pages still have selectable text.
Combining a long report with a short appendix.
Two files
40-page report (165,897 bytes) + 12-page appendix (50,563 bytes)
One file
52 pages, 216,125 bytes, 35 ms
216,460 bytes went in and 216,125 came out, a 0.2% difference. Fonts embedded in both documents are not deduplicated by this process, so a merged file is very close to the sum of its parts rather than smaller than it.
Real files, real numbers, and the machine they were run on.
| Scenario | Input | Output | Time | Notes |
|---|---|---|---|---|
| 40-page and 12-page text PDFs into one | 216,460 bytes | 216,125 bytes | 35 ms | 52 pages, text still selectable |
| 12-page scan and a 12-page text PDF | 10,374,682 bytes | 10,374,095 bytes | 32 ms | images untouched at original resolution |
How this was measured: Measured by driving this page in headless Chrome 152 on an Apple M4 Pro running macOS 15.5, on 2026-09-17. The fixture is handed to the page's own file input, the page's own button is pressed, and the figure recorded is the size of the file the page produces. Timings come from a warm page on a fast machine with no network involved, so treat them as a floor rather than a typical result.
A countersigned agreement usually arrives as the scan of a signature page plus the typed original. Combining them on your own machine means the signed copy is never handed to a third party whose retention policy you have not read.
Letters, test results and discharge summaries routinely need to travel as one document. These are exactly the files that internal policy forbids putting through a consumer upload service, which is usually why people end up doing it by hand.
Portals that accept a single attachment force you to staple a CV, transcripts and certificates together. The order matters to whoever reads it, which is what the drag-to-reorder list is for.
A month of receipts scanned individually becomes one submission. Nothing here is per-file limited, so a fifty-receipt claim goes through in the same pass as a two-receipt one.
Plenty of workplaces block the well-known PDF sites outright. A page that does the work locally keeps working when the upload-based alternative is unreachable, and does not need an exception raising to use it.
Once this page has loaded it keeps merging with the network off, because the code that does the job is already in your browser. Useful on a plane, and a straightforward way to satisfy yourself nothing is being sent.
Structural facts about PDFs, not shortcomings of this page, worth knowing before you send the result on.
This is the one to know about. A document with form fields comes back looking identical, and the fields no longer work: the values and the field definitions are dropped while their appearance survives as flat page content. Merging a fillable form produces a picture of that form. Check before sending anything on.
The navigation tree that some readers show down the side is stored on the document rather than on its pages, so copying pages into a new document leaves it behind. A 400-page manual will merge correctly and arrive with no chapter list.
Merging is a structural operation, so the output is approximately the sum of the inputs. If the combined document is too large to email, that is a separate job.
You can arrange the documents, not the pages inside them. To interleave pages, or to move page nine ahead of page two, merge first and then rearrange.
Open your browser's developer tools, switch to the Network tab, and merge something. You will see this page's own assets load and then nothing at all while the work happens. The stronger test is to disconnect from the network entirely once the page has loaded and merge anyway. It still works, which is not something a tool that needs a server can do.
No. Scanned pages are images inside the PDF, and those image objects are copied across byte for byte. Nothing is decoded, resampled or re-encoded, so a 300 dpi scan is still a 300 dpi scan afterwards, at the same file size it contributed going in.
Because nothing was thrown away. A merge copies the objects that describe your pages into one container and rebuilds the index; it does not recompress images or subset fonts. Two documents totalling 216,460 bytes produced 216,125 bytes. Expect the sum, not a saving.
Yes. Each source document brings its own embedded font programs, and they are copied independently rather than matched up and shared. For two documents using the same typeface this adds a little weight, which is part of why the output lands slightly above the raw sum before the cross-reference saving pulls it back.
Not while the encryption dictionary is still there. Page objects cannot be copied until they can be read. When the secret is one you already have, Unlock PDF writes a clear copy, and that copy is what you drop into the list here.
None that this page imposes, because there is no server tier behind it to enforce one. The real ceiling is your device's memory: every document is held in RAM while the new one is assembled, so a laptop copes with far more than a phone does. A 10 MB merge is unremarkable.
Exactly the order the files appear in the list, top to bottom, with each document's own pages kept in sequence. Dragging a row changes the output; the filenames themselves are ignored, so you do not need to rename anything to force an order.
Page-level annotations such as highlights, sticky notes and link areas are attached to the pages themselves, so they are carried across. Anything stored at document level rather than page level. The bookmark tree, form field definitions, is not.
Not here. This page reads PDF page trees, which images and Word documents do not have. Convert them first and then merge the results.
Yes, with the caveat that the work happens on the device. Phones have less memory to spare and the merge runs on the main thread, so a large scan-heavy job will make the tab unresponsive for longer than the same job on a laptop.
No. The documents exist as objects in the page's memory and the merged result as a temporary blob, all of which the browser discards when the tab closes. There is no cache to clear and no history to delete because nothing was written anywhere.
Yes, once the page has loaded. Everything that does the merging is already in your browser at that point, so you can switch off the network and carry on. Reloading while offline depends on the browser still holding the page in its cache.
There is no limit imposed by us, because there is no server tier to enforce one. The real constraint is your device's available memory, since the whole file is held in RAM while it is processed. Very large files are slower on phones than on a laptop.
No. There is no sign-up, no email, and no usage tracking tied to an identity. The tool works the first time you open it.