Merge PDF

The jobs people actually use this for

Merging is rarely the point in itself. Almost always it is the last step before something has to be sent somewhere as a single file, and the receiving end has rules.

A few that come up constantly:

• Expense claims. The finance system accepts one attachment, but you have eleven receipts photographed as separate PDFs from a phone scanning app.
• Contracts with appendices. The signature page, the schedule, and the pricing annex arrive as three files from three different people, and the client wants one document to initial.
• Rental or visa applications. Requirements list "passport, bank statement, proof of address, in one file, in this order" — the order matters, and getting it wrong means the application is rejected.
• Scanned book chapters or lecture handouts. Each chapter came off a photocopier as its own PDF and you want a single searchable-by-eye file to read on a tablet.
• Court or insurance submissions, where exhibits are numbered and the numbering has to match a written index. This is the case where getting the sequence wrong is expensive, so build the order from the index rather than from memory.

What actually happens when you press Merge

When you select a file, the browser hands the page a File object that points at the file on your disk. It is not a copy sitting in a server folder — it is a handle. The page reads its bytes into memory only when it needs them, using the standard ArrayBuffer API that every browser implements.

The merging itself is done by pdf-lib, a JavaScript library that runs in the tab. It parses each source document, copies each page object into a new document, and serialises the result back to bytes. Those bytes become a Blob, and the Blob gets a temporary in-memory URL that the download link points at. When you close the tab, that URL and the memory behind it are gone.

If you want to satisfy yourself that this is true rather than take our word for it, open your browser developer tools, switch to the Network panel, and press Merge. You will see the request for the pdf-lib code the first time, and then nothing. No request carries your document. The same test works on any tool on this site, and it is the only claim here worth checking yourself.

What is preserved, and what quietly is not

This is the part most merge tools do not tell you, and it is the part that occasionally ruins someone's afternoon.

• Page size and orientation are preserved per page. Mixing A4 and US Letter in one document works; each page keeps its own dimensions rather than being scaled to match.
• Visual quality is untouched. Pages are copied as objects, not re-rendered as images, so text stays selectable and there is no generation loss. The merged file will be roughly the sum of the inputs, which is why merging is not a way to make a document smaller.
• Bookmarks and the outline tree are lost. The merged document will not have the navigation sidebar the sources had.
• Form fields do not survive. Fillable fields become flattened or disappear, and any data a recipient typed is not carried across. Fill forms before uploading, or merge first and fill afterwards — not both.
• Digital signatures are broken by definition. A signature covers the exact bytes of the document it was applied to. Adding a page changes those bytes, so the signature will report as invalid even though nothing was forged. If a signed contract must stay verifiable, do not merge it — send the parts separately.
• Attachments embedded inside a PDF are dropped.

How to build the right order quickly

The single most common mistake is adding files in the order the file manager happened to list them, then discovering halfway through the merged document that page 4 is where page 9 should be. Merging again with the order fixed takes ten seconds, so this is an annoyance rather than a disaster — but there is a cheap way to avoid it.

If the document has a written index, put the index next to you and add files in index order, one at a time, checking the list as you go. If it does not, rename the source files so alphabetical order matches the order you want, then add them all at once — sorting happens in your operating system, where it is fast, instead of in a list where it is manual. Two-digit numbers (01, 02, … 10) keep the order correct past nine files, which single digits do not.

When merging is the wrong tool

Three situations where you want something else:

• The merged file is too big to email. Merging does not compress, so you need the PDF compressor, which re-encodes the embedded images. Compress first, merge second — doing it the other way round means compressing one large file instead of several small ones, which is slower and gives a worse result.
• A source file is password-protected. It cannot be merged while encrypted, and this tool will tell you which file is the problem instead of silently producing blank pages. Open it once in a reader, enter the password, and export a decrypted copy.
• You are combining hundreds of large files. Everything here runs in the memory the browser gives the tab, which is a fraction of your total RAM. A handful of files is nothing; a project of five hundred 50 MB scans is better done in batches, merged in groups of twenty or so and then combined.