01What local PDF merging means
Swiftly reads each PDF in your browser, copies the selected pages into a new document with pdf-lib, checks the finished page count and creates the download on the same device. There is no merge endpoint and no document-processing API key. Your PDF bytes are not sent to Swiftly or a third-party merger.
“No upload” describes document processing, not every network request. The page and its dynamically loaded PDF library must be available before files can be inspected. A first visit or cleared cache can therefore require a connection even though the documents themselves remain local.
No account, watermark, visible advertisement or promotional page is added to the merged PDF.
02Set file order and choose exact pages
The list runs from first to last. Drag a row to a new position, or use the up and down buttons for a keyboard- and touch-friendly alternative. You can add more PDFs before merging, remove a file, or clear the batch. A single PDF is also accepted when you only need to select, reorder, duplicate or rotate its pages.
Open “Pages & rotation” on any file to use a custom selection. Page numbers are one-based: 1 is the first page. Commas control the order, ranges such as 1-4 include both endpoints, duplicates repeat a page, and a descending range such as 9-7 inserts pages 9, 8 and 7. An open range such as 20- continues through the final page. For example, 3, 1-2, 3 produces page 3, pages 1 and 2, then another copy of page 3.
- All pages is the default for every added file.
- Custom selections preserve the order you type rather than silently sorting it.
- The planned output count appears before you merge.
- The assembled page count is checked against the plan before the PDF is created for download.
03Rotation, page sizes and visual quality
Rotation applies only to the pages selected from that file and is added to any rotation already stored on the source page. Choose 90° clockwise, 180° or 90° counterclockwise. Different source page sizes remain different; the tool does not stretch Letter, A4, landscape or custom pages onto a forced canvas.
Merging copies PDF page objects and content streams rather than rasterizing every page. It does not re-encode photographs as JPG or reduce their resolution. The output container is rewritten, so its byte size can be larger or smaller than the sum of the inputs, but a smaller file is not presented as proof of better compression or quality.
This version uses page numbers and ranges, not rendered thumbnail editing. Check a source’s page numbering in a PDF reader before entering a complex selection.
04Duplex blank pages and optional page numbers
“Start each source on a new sheet” is intended for double-sided printing. When a source contributes an odd number of selected pages, the tool inserts a blank page after it so the next source begins on a fresh sheet. The blank uses the preceding page’s dimensions. No trailing blank is added after the final source.
Footer numbering is off by default. When enabled, every output page—including an inserted blank—receives a centered “current / total” label near the bottom. Existing footer content is not detected, so the label can overlap text, signatures, stamps or trim marks. Review every numbered page before printing or publishing.
05No fixed caps, passwords and browser memory
Swiftly does not impose a fixed file-count, per-file-size, batch-size, input-page, output-page or output-byte cap. This is not a promise of infinite processing: every browser and device has finite memory and execution limits, and a complex PDF can require much more memory than its file size suggests.
The engine stores ranges as compact segments, copies pages in small work batches and avoids a second full parse of the finished output solely for validation. These choices reduce avoidable allocations but cannot eliminate the memory required for input bytes, copied resources and the final PDF. Password-protected and encrypted PDFs are rejected rather than silently unlocked. Corrupt files receive a readable error. If a large merge exhausts browser memory, close heavy tabs or split the job into smaller merges.
- No fixed Swiftly cap on the number of PDFs or input/output bytes.
- No fixed Swiftly cap on source pages or merged output pages.
- Open-ended ranges such as 20- continue to the source’s final page.
- PDF input only; practical capacity depends on the browser, device and file complexity.
06Bookmarks, forms, signatures and final checks
Copying pages is not the same as preserving every document-level feature. Existing page content and many annotations can survive, but existing outline trees, embedded attachments, interactive form structure, JavaScript and other catalog-level features are not guaranteed. The optional bookmark setting creates a new flat outline with one filename entry per source; it does not reconstruct the original bookmark trees. Any digital signature must be treated as invalidated when a signed document is modified or its pages are copied into a new file.
The output receives a new title and Swiftly producer/creator metadata rather than carrying forward every source metadata field. Keep the original documents for provenance and archival needs. Before sharing, open the merged PDF and verify page order, custom ranges, rotation, blank-page breaks, page-number placement, links and form behavior.
Do not use a merged copy to represent an unchanged signed original. Keep the signed source and ask the recipient what form of document they require.