Drop your PDFs here
Drag PDFs or pictures onto this page, or press the button. Every page shows up as a thumbnail you can drag, rotate and delete. Then pick a tool. Nothing is uploaded.
pdf [merge|split|compress|images|topdf|numbers|watermark|crop|info] fileDrag PDFs or pictures onto this page, or press the button. Every page shows up as a thumbnail you can drag, rotate and delete. Then pick a tool. Nothing is uploaded.
Your files are opened and changed by your own browser. The page loads the open-source PDF code it works with from public CDNs, and that is the only thing that travels.
Press Add files or drop PDFs and pictures anywhere on the page. Every page shows up as a thumbnail. Nothing is uploaded, and a few sample files let you try everything first.
Drag pages into order, click to select them (Shift selects a range), then rotate, duplicate or delete. Pages from different files sit side by side, and Ctrl+Z undoes a mistake.
Choose Merge, Split, Compress, PDF to images, Images to PDF, Page numbers, Watermark, Crop or Text & info. Its options appear under the tool buttons.
Say how to split, how small the file should get or the size you need, which resolution, where the numbers go or what the watermark says. The red frames in Crop show what stays on the page.
Press the big button at the bottom of the panel. Download each file or a ZIP, or choose Continue with this file to run another tool on the result without opening it again.
Most online PDF sites have one page per job: a merge page, a split page, a compress page. This page puts them together. You add your files once and every page appears as a thumbnail, so you can see what you are working on, drag pages into order, turn a sideways scan, throw away a blank page, and only then press the button for what you need. The result can be fed straight into the next tool, which is how you get from three scans to one numbered, compressed PDF in a minute. The nine tools sit in the rail on the left, also under the keys 1 to 9. The page tools (select all, rotate, duplicate, delete) stand next to the thumbnails, and the status line under the pages shows the files, the number of pages, the selection, the total size and what the big button will make; click a part to go there. For the keyboard there is a command prompt: press Ctrl+K (⌘+K on a Mac) or /, type compress, png, page numbers, select 2-5 or ranges 1-3, 5 and press Enter.
It is also private. The files are read, changed and saved by your browser. They are not sent to a server of mine or anyone else's, there is no account and there is no watermark on the result. That is the difference with most free PDF websites, which upload your documents, limit the number of files per day and stamp a logo on the output.
PDFs and pictures side by side, drag to reorder, rotate, duplicate and delete pages, mix the scans of a two-sided document, undo and redo.
By page ranges, every N pages, one file per page, selected pages, a maximum file size or the document's bookmarks.
Four quality levels, or a target size such as 1 MB. Pictures are recompressed, text and drawings stay sharp and selectable.
PDF to JPG, PNG or WebP at up to 300 dpi, and pictures to PDF with A4, Letter or picture-sized pages.
Page numbers in six positions and five formats, and a text or picture watermark with opacity, angle and a repeating layout.
Trim margins, fit pages onto a paper size, edit or remove the document properties and extract the text.
Add two or more PDFs and they join into one list of pages, coloured by file, in the order you added them. Drag a page, or a whole selection, to a new place; the line between the pages shows where it will land. Use Sort files A–Z to put numbered files in order, Reverse order for a document that was scanned from the back, and Mix two scans for the common office problem of a two-sided document scanned as a pile of fronts followed by a pile of backs: the second half is woven in between the pages of the first, with a variant for backs that came out in reverse.
Merging copies each page as it is. The text stays text, links and bookmarks on the pages remain and the quality does not change. Pictures you add become pages too, so a contract, a signature scan and a photo of a receipt can end up in one file. Form fields do not survive a merge as fields, because the pages are combined into a new document. To fill in a form, use the PDF form filler before or after.
Splitting follows the page list you see, so you can rearrange first. Page ranges take a list like 1-3, 5, 8- and make one file per range; switch on keep all ranges in one file to extract those pages instead. Every N pages cuts a long scan into equal parts, each page makes a file per page, and selected pages saves the pages you clicked as a new PDF. Maximum size packs whole pages into files that stay under a limit, for an e-mail server that refuses attachments above 10 MB, and bookmarks cuts at the chapters of a document that has them. Several files arrive together in a ZIP, with a download button for each one.
A PDF is large when it contains pictures: a scanner or a phone camera stores a full-size photo of every sheet. The compressor opens the file, finds those pictures and saves them again as smaller JPEGs, then writes the PDF back. Text, fonts, vector drawings, bookmarks and links are not touched, so a report with a few photos shrinks a lot and stays sharp, while a file with only text is already small and the tool says so instead of pretending. A result that would be bigger than your original is never handed out.
| Level | Longest side of a picture | JPEG quality | Good for |
|---|---|---|---|
| Light | 2000 px | 74% | Documents you still print or zoom into |
| Balanced | 1400 px | 58% | E-mail attachments and portal uploads (recommended) |
| Strong | 1000 px | 42% | Reading on screen, much smaller |
| Extreme | 640 px | 28% | The smallest file; pictures turn soft |
With Target size you name the limit instead, for example 2 MB or 500 KB, and the page tries stronger and stronger settings until the file fits, keeping the best quality that does. If the pictures alone cannot get there it turns the pages into pictures as a last step and tells you so. Turn pages into images forces that route for scans and for files that resist compression; the text is then no longer selectable. Grayscale drops the colour, and each file separately compresses every file on its own.
A PDF page is measured in points, 72 to the inch, so the resolution you choose decides how many pixels each page gets: at 150 dpi an A4 page becomes about 1240 × 1754 pixels, at 300 dpi about 2480 × 3508. Pick PNG for sharp text and drawings, JPG for photos and small files, WebP when the picture goes on a website. Several pages come as a ZIP.
For the other direction add JPG, PNG, WebP or GIF pictures and choose a page size. A4 and US Letter centre each picture on a page with a margin, in portrait or landscape to suit, while same as the picture makes every page exactly as big as its picture. JPG and PNG files are embedded as they are, without being re-encoded, so a photo does not lose quality in the conversion; a lower quality setting re-encodes them to save space. Photos taken on a phone often carry a rotation flag, which is applied so they come out the right way up.
Page numbers can sit at any of six places on the page, as 1, Page 1, 1 / 12, Page 1 of 12 or – 1 –, starting at any number, with the size, margin and colour you want. Skip the first page for a cover, or number only the selected pages. The numbers follow the page as it is shown, so they stay upright on a page that is turned sideways.
A watermark can be text or a picture, drawn with the opacity and angle you set, once in the middle or repeated over the page. Text is turned into a picture first, so every script your browser can display works, including accents, Cyrillic, Greek and Chinese. Watermarks lie on top of the page; they discourage copying but do not make a file tamper-proof.
Type how many millimetres to trim from each edge, or pick a preset, and the pages show a red frame around what will remain. Fit onto paper then scales every page, without distortion, onto A4, A5, A3, US Letter or US Legal, so a mixed pile of scans ends up uniform. Cropping sets the visible area of the page, which is how PDF viewers do it too: the cut-off part is hidden and not erased, so it is not a way to remove confidential information. For that, flatten the pages to images first with the turn pages into images option in Compress.
Every PDF carries properties such as title, author, subject, keywords, the program that made it and dates. Fill in the fields to set them, or switch on remove all properties to send a file without a trace of who made it and with what; the list on the page shows what each file contains now. Extract text saves the text of the pages as a .txt file, or copies it, for the pages that have real text. A scan is only a picture of text, so it needs OCR, which this page does not do and will tell you.
A PDF with an open password asks for it when you add it, and the password stays in your browser. Files that carry restrictions on editing or copying cannot be copied page by page without breaking the protection, so their pages are rebuilt as pictures at a good resolution; the file list marks them locked. The result looks the same and can be merged, split and stamped like any other file, but its text is no longer selectable.
Your PDFs and pictures are read into the memory of this tab, worked on there and saved from there. They are not uploaded. What the page does download, the first time it needs it, is open-source code: pdf-lib from unpkg (with jsDelivr as backup), PDF.js and its worker from cdnjs (jsDelivr as backup), and JSZip from cdnjs when a ZIP is made. PDF.js may also fetch standard fonts or character maps from jsDelivr for a PDF that needs them. Those servers see a request for a library file and your IP address, as any website does, and nothing about your documents. Page statistics come from SimpleAnalytics, which does not use cookies. The page remembers your last tool, the size of the thumbnails and your option choices in the browser's local storage under the keys jb-pdf-ui and jb-theme (light or dark), never the files or their names.
Everything depends on the memory of your device. A phone handles files of tens of megabytes, a desktop browser several hundred; above 600 MB in one go the page asks you to confirm. Thumbnails are drawn only for the pages on screen to keep memory down, and very large pages are drawn at a lower resolution when the browser's canvas limit would otherwise be exceeded. Interactive form fields, digital signatures and tagged accessibility structure are not preserved when pages are copied into a new file. If a result looks different from what you expected, press Undo, change the options and run it again: your original files are never modified.
To fill in a form or sign a PDF use the PDF form filler; to translate a whole document use the PDF translator. Text pulled out of a PDF often needs tidying, which the text cleaner does. Find more in the full list of tools.