PDFMerge

PDFMerge is a free, privacy-first web app for merging PDFs and images into a single document. All processing runs entirely in your browser: nothing is uploaded to any server.

Live at pdfmerge.me

Key Features

File Handling

• PDF and Image Merging: Combine PDFs with JPG, PNG, WEBP, and GIF images
• Drag and Drop: Add files by dropping them anywhere on the page, or reorder cards by dragging. Keyboard users can choose files with Choose files and move a focused card with Alt+Arrow keys or Alt+Home/End
• Sort Files: Sort file cards once by natural filename order (A-Z or Z-A) or modification date (oldest or newest first). Blank pages stay in their slots
• PDF Page Selection: Specify page ranges (1-5), individual pages (1,3,7), or a mix (1-3,5,8) per PDF. Pages merge in the order typed, so 3,1,2 reorders and 1,1 repeats a page. Invalid tokens or page numbers block the merge and show an error under that card
• Cropped PDF Pages: The visible CropBox region fits within the selected paper size. Letterbox margins stay blank, and page text and watermarks can use the full output page.
• Duplicate Files: Add the same file more than once, for example to take different page ranges from one PDF
• Blank Page Insertion: Add blank separator pages to the output
• Individual File Removal: Delete specific files with per-card remove buttons

Image Layout Options

Each image can be placed on its page in one of three modes:

• Default Margins: Scaled within the image page margins selected in Margins. Default is 54 pt on every side (about 19 mm); choose Narrow (12.7 mm), Normal (25.4 mm), None, or Custom per-side values from 0 to 50 mm in 0.5 mm steps. Cover and Fit still use the full page.
• Cover: Cropped to fill the full page
• Fit: Scaled to fit with letterbox background

Transparent image areas become white in the output PDF in every mode.

Use the Toggle Image Rotation button to rotate the image content 90, 180, or 270 degrees independently of the layout mode.

Output Options

• Paper Sizes: A0, A1, A2, A3, A4, Letter, Legal, Tabloid
• Page Orientation: Toggle any image page between portrait and landscape
• Output Page Numbers: Number every output page, including blank and duplex padding pages. Choose bottom centre, bottom right, or top right; a preset or custom template of up to 100 characters; and a starting number of 1 or greater. Templates use {n} for the current number, {total} for total output pages, {file} for the source name without its extension, and {date} or {time} for the local time captured when merging starts. Use {n:3} or {total:3} for zero padding (width 1–6), and {{ or }} for literal braces. Blank-page cards have an empty file name; duplex padding uses the preceding card's name.
• Header and Footer Text: Add text on every output page, including blank and duplex padding pages. Each has a template of up to 100 characters and left, centre, or right alignment. Header defaults to centre; footer defaults to left. They use the same placeholders as page numbers, but plain text such as CONFIDENTIAL is valid. All three text options share the “Numbering starts at” value for {n}. If page numbers use the same edge and alignment as a header or footer, choose another alignment before merging.
• Skip First Page: Page numbers, header text, and footer text each have this option. The first output page still counts toward {n} and {total}; if skipped, page 2 uses the next number.
• Text Colour: Choose a separate colour for page numbers, header text, and footer text; each defaults to grey.
• Text Style: Page numbers, header text, footer text, and the text watermark each have bold, italic, and underline toggles. Text defaults to normal weight; the text watermark defaults to bold. Italic is a slanted Roboto, as no italic font file is bundled.
• Text Background and Margin: Each text option can add a solid background box with its own colour. Set its top or bottom and side margin from 0 to 50 mm in 0.5 mm steps; the default is 8.5 mm.
• Margins: Header and footer margins now live in Margins. They apply only when their text is enabled. Image page margins are saved across sessions and shown in the live preview.
• Duplex Padding: Add a matching blank page after each card that produces an odd number of pages, except the final card
• Image and File Details: Add EXIF details, a file hash, or an image number through placeholders in any page number, header, or footer template: {taken} (capture date and time, accepting the same tokens as {date}; default YYYY-MM-DD HH:mm), {gps} (latitude, longitude), {hash} (SHA-256 of the source file; {hash:8} keeps the first 4 to 64 characters), and {img} (image number counting image files only; {img:3} pads it). {hash} works for every source file; {taken}, {gps}, and {img} print nothing on pages without that data, such as PDF or blank pages
• Custom Filename: User-defined output filename
• Remove Metadata: On by default. Clears the merged PDF's title, author, subject, keywords, creator, producer, creation and modification dates, and catalog XMP metadata. Turn it off to keep the existing output behavior.

Date and time default to {date} = YYYY-MM-DD and {time} = HH:mm. Either placeholder accepts the same tokens after its first colon, as in {date:YYYY-MM-DD HH:mm}. Wrap literal letters in brackets, for example {date:[Day] D}.

Token	Output
YYYY, YY	Four-digit or two-digit year
MMMM, MMM	English month name, full or abbreviated
MM, M	Month, padded or unpadded
DD, D	Day of month, padded or unpadded
dddd, ddd	English weekday name, full or abbreviated
HH, H	24-hour clock, padded or unpadded
hh, h	12-hour clock, padded or unpadded
mm, ss	Padded minutes or seconds
A	AM or PM

Examples: Revised Page {n} of {total} - {date:D MMM YYYY}, {file} {date:YYYY-MM-DD} {time:hh:mm A}, and Image {img} taken {taken:D MMM YYYY h:mm A} GPS {gps}.

Watermarking

• Text watermark with colour, opacity, rotation, and scale
• Image watermark from a PNG (transparency kept) or JPG: nine positions (corners, edge centres, centre), width as a percentage of the page (5–100%), and margin in mm for non-tiled edge positions, plus the shared opacity and rotation controls. The image stays in memory for the current tab only; it is never uploaded or saved, so choose it again after reloading.
• Single watermark or tiled pattern across every page

Live Preview

• Shows output pages with every output option applied, built by the same code path as the merge, so it matches the downloaded file
• All preview pages share one scale, so portrait and landscape pages keep their true relative sizes
• Updates shortly after any change; invalid settings dim the last preview until fixed
• Desktop shows it beside the settings; on small screens it is a collapsible section above them
• Enlarge a page to inspect details, zoom from fit to 400%, pan, and move between pages

Privacy and Performance

• No server uploads: all PDF and image processing happens locally in your browser; your files are never uploaded
• Self-hosted assets: all fonts, scripts, and assets are served from the same origin (no third-party CDNs)
• Security headers (CSP with frame-ancestors 'none', X-Content-Type-Options, Referrer-Policy, Permissions-Policy) set through the Cloudflare Pages _headers file
• The only external request is to Cloudflare Web Analytics (cloudflareinsights.com), a cookieless, privacy-preserving page-view counter that collects no personal data and no file content
• Cancel mid-merge at any time
• Progress bar with 60-second timeout protection
• Dark and light mode, persisted across sessions
• Option toggles and watermark settings persisted to localStorage across merges and sessions; a notice lists the sections that differ from the defaults, with a Reset to defaults button
• Settings below the file cards use eight collapsible sections; each section remembers whether it was open
• Cancelling the save dialog keeps the file list so the merge can be run again
• The download-again link stays available until you change files, start another merge, or reset
• Notifications: Dismissible toasts confirm file additions, removals, sorting, clearing, blank pages, and merges
• Accessibility: Visible keyboard focus, 44px touch targets, screen-reader status for long operations, and reduced-motion support

Browser Compatibility

• Optimal: Google Chrome and other current Chromium-based browsers
• Good: Firefox, Safari (including iOS and iPadOS), and Microsoft Edge (latest versions)
• Requirement: A modern browser with ES module, File API, and Web Crypto support

Development

Install the pinned development tools with npm ci, then run:

• npm run lint to lint first-party JavaScript and tests
• npm test to run the regression test suite

Versioning

The version lives in package.json. CI bumps the patch version on every push to main that changes app files, after tests and lint pass. Pushes containing only Markdown, docs/, or LICENCE.txt changes are ignored. Local assets and module URLs carry ?v=<version> to prevent stale CSS and JavaScript after deploys. Run npm run version:sync after editing to sync the footer and asset URLs; never hand-edit ?v=. CI commits these synced files with the version bump so Cloudflare Pages deploys them. For a manual minor or major bump, use the Tests workflow's Run workflow button on main and select the bump type.

Known Limitations

PDFMerge runs client-side, so performance is limited by your device and browser. Image files are capped at 50 MB due to browser memory constraints when resizing via the Canvas API. For very large images, reduce size or resolution before uploading. Chrome generally handles larger batches better than Firefox.

Technical Notes

• Architecture: Static single-page app with vanilla JavaScript ES modules: no build step, no framework, no server
• PDF Engine: pdf-lib for PDF creation and manipulation; PDF.js for thumbnail rendering
• Font Handling: UI fonts (Roboto, JetBrains Mono) are self-hosted in fonts/web/. Roboto is also embedded in output PDFs via fontkit using files in fonts/
• EXIF Extraction: exif-js reads metadata from JPEG files before embedding
• Security: Content Security Policy restricts default-src to 'self'; the only permitted external origin is Cloudflare Web Analytics (script-src/connect-src). No third-party scripts, fonts, or trackers are loaded

Support

If pdfmerge is useful to you, please consider starring the repository to help others discover it.

Credits and Third-Party Licensing

• pdf-lib v1.17.1 by Andrew Dillon | MIT License
• PDF.js v4.10.38 by Mozilla | Apache License 2.0
• exif-js v2.3.0 | MIT License
• fontkit v1.1.1 | MIT License
• Roboto Font by Christian Robertson | Apache License 2.0
• JetBrains Mono by JetBrains | SIL Open Font License 1.1
• Lucide v1.61.0 | ISC licence. Icons are self-hosted in images/icons.svg; no external icon requests