Merging PDFs sounds like one task but splits into two very different situations: files you're happy to upload to a website, and files you absolutely are not — contracts, medical records, financial statements. Pick the wrong route for the second category and you've handed a third party your sensitive documents. Here's every method, what it does to your pages, and when to use it.
Nothing dramatic, which is the good news. A PDF is a container of page objects referenced through a cross-reference table; a merger builds a new container and copies each input's pages into it, in order. Pages aren't re-rendered or recompressed — text stays selectable, scanned pages stay scans, embedded fonts stay embedded, and each page keeps its own dimensions (mixing letter-size and A4 in one file is fine). The output size lands close to the sum of the inputs, with small deltas from the new cross-reference structures and deduplicated shared objects.
Three things merging does not do: it doesn't build a table of contents or carry over source bookmarks in most tools, it doesn't renumber baked-in page footers (those are page content), and it doesn't preserve digital signatures — a signature covers the original file's exact bytes, so any merge invalidates it, no matter whose tool did the merging.
| Method | Upload? | Cost | Pick it when… |
|---|---|---|---|
| Browser merger (no upload) | No | Free | You want one-click speed with desktop-level privacy |
| Adobe Acrobat desktop | No | Subscription | You merge daily and need bookmarks/portfolios |
| macOS Preview | No | Free | You're on a Mac — it's already installed |
| qpdf / pdftk CLI | No | Free | You merge the same files repeatedly or in pipelines |
| Online upload sites | Yes | Freemium | Non-sensitive files and no local option exists |
Open the PDF Merge tool, drag your files onto the box (or click to pick them), arrange the rows in the order you want the pages to appear, name the output, and click Merge. The page count of every file is shown as it loads, so you can sanity-check before committing — three files showing 12, 8, and 5 pages should merge into 25. The tool runs on the pdf-lib engine (MIT-licensed, maintained as @cantoo/pdf-lib) entirely in your browser's memory: open the network tab and you'll see zero requests during the merge itself. That's the whole difference from upload-based services — there's no copy of your document on anyone's server to retain, breach, or train on.
Preview is the most underrated merger because every Mac has it. Open one PDF in Preview, show the sidebar (View → Thumbnails), then drag the second PDF's icon into the sidebar below the last thumbnail of the first — the insertion point matters, since that's where its pages land. Drag thumbnails within the sidebar to reorder pages across both files, then File → Export as PDF (or Save, in recent versions) to write the combined file. For quick two-file merges on a Mac it's genuinely the fastest path; its weakness is large batches and no scripting.
Two one-liners cover almost everything. With qpdf: qpdf --empty --pages a.pdf b.pdf c.pdf -- merged.pdf — empty document, copy all pages of each listed file in order. Want only part of a file? Add ranges: qpdf --empty --pages a.pdf 1-3 b.pdf 5-9 -- merged.pdf. With pdftk: pdftk a.pdf b.pdf cat output merged.pdf, using the same cat concatenation model. Both are free, run offline, and are what most document pipelines use under the hood. The browser tool is doing conceptually the same operation — copying page trees into a fresh document — without the terminal.
Merging is the wrong verb when you need pages 2–4 of one document combined with another: that's extract-then-merge. The PDF Split tool handles the cutting side — ranges, fixed-size chunks, or one-file-per-page bursts — and pairs naturally with merge for assembling custom packets (pull pages 1–3 of the contract, then merge with the full report). The same qpdf range syntax above does it in a terminal.
Drag, order, merge, download. No upload, no watermark, no sign-up, works offline once loaded.
Merge PDF Files →Merge method = trust boundary. Browser-based or CLI keeps files on your machine; upload sites don't; Acrobat costs money for extras most merges never need. The operation itself is lossless page concatenation in every case — so choose on privacy and workflow, not on output quality, and sign documents only after the final merge.
Use a browser-based merger like ToolAspect's PDF Merge: open the page, drag your files in, order them, and click merge. The work happens locally with the pdf-lib engine, so the files never leave your device and it works even if you go offline after the page loads. Command-line users get the same privacy with qpdf or pdftk.
Pages appear in file order, file by file: all pages of the first document, then all of the second, and so on. Merging files of 12, 8, and 5 pages yields pages 1-12 from the first, 13-20 from the second, and 21-25 from the third — 25 pages total. Reorder the files before merging to change the map; reorder pages inside a file with a split-then-merge workflow.
No quality change: pages are copied rather than re-rendered, so text, vectors, image resolution, and embedded fonts survive intact. Size lands near the sum of the inputs — small deviations come from the new document's cross-reference tables and the deduplication of shared objects, not compression.
Digital signatures cryptographically cover the exact bytes of the original file. Any later change — merging, page edits, even re-saving — alters those bytes and invalidates the signature. Every tool has this behavior, including Acrobat. Sign the final merged document, not the pieces.