Drop your document into the zone below. Every page appears as a tile you can turn, so rotate one page, a run of pages, or the whole file, then save the rotation permanently so it opens upright in every reader you send it to. Nothing is re-rasterized, so the text stays selectable, the links keep working and the file size does not move. Your document is never uploaded.
Your files are read by JavaScript running on your own machine. There is no upload, no queue, and no copy on our server to delete later.
Drag a card to reorder it, or use the arrows under each card. The order here is the order in the output. Each card is a preview of the finished page, so what you see is what the PDF will contain.
Esc close ← → move R rotate
Starting
This is your processor doing the work, so the time it takes depends on your machine and not on how many other people are on the site.
The tool above is ready to use. The guide below explains the thing that catches almost everybody: rotating a page in a PDF reader usually does not save. It explains why a scanner produces a stack where eighty pages are sideways and twelve are not, what the rotation is actually stored as inside the file, and why rotating a PDF here costs nothing in quality while re-exporting from a print dialog costs you the text layer.
About a minute, nothing to install, and no account needed for your first run. This PDF rotation tool executes inside the page you are reading, so all four steps happen on your own machine.
Click the drop zone above or drag the file anywhere onto this page. Every page is read out of the document and drawn as a tile, so you can see the whole thing at once instead of scrolling a viewer.
Each tile has an arrow to turn it left and an arrow to turn it right. Press Stand every page upright and only the landscape pages move, which fixes a mixed stack in one click. The badge under each tile shows the turn it is carrying.
Click any tile to open that page full screen. A thumbnail is enough to tell portrait from landscape and nowhere near enough to tell upside down from the right way up on a page of dense text, so check at least one before you run it.
Press Rotate PDF. The rotation is written into each page dictionary and the file comes back immediately. Open it, and it opens upright in every reader, on every device, for everybody you send it to.
Keyboard works throughout. Tab to a page, press R to turn it right and Shift R to turn it left, and Enter to open it full size. On a phone every control is a real button and nothing depends on a gesture, because a rotation gesture on a grid of forty tiles is a rotation nobody can aim.
This is the single most common reason people end up on a page like this one. You opened the file, you turned the page the right way up, you closed it, you emailed it, and the recipient told you it is still sideways. Nothing broke. You did nothing wrong: most PDF readers treat rotation as a view setting and never write it into the document.
The distinction is real, and it decides which tool you need:
/Rotate, an integer that is 0, 90, 180 or 270. It instructs every viewer to turn that page clockwise by that many degrees before showing it. Change that number and the page is rotated for everyone, forever.This tool writes the second kind. It changes the number in the page dictionary and touches nothing else in the file. The content streams are byte identical to the ones you dropped in, which is why the operation finishes as fast as your disk can hand over the bytes even on a two hundred page document.
A quick way to tell which kind you are looking at. Rotate the page in your reader, save the file, close it, then open it again from your downloads folder instead of from the recent files list. If it is sideways again, you were given a view rotation. Almost every reader behaves this way and almost none of them tell you.
There are two ways to rotate a page in a PDF and they are not remotely equivalent. Understanding the difference is the difference between a 4 MB file that is still searchable and a 40 MB file that is a stack of photographs of your own document.
A PDF page is a dictionary of properties, and one of them is /Rotate. Set it to 90 and every conforming reader displays that page turned a quarter turn clockwise. The page content, the fonts, the images, the text positions and the annotations are all untouched, because none of them ever needed to change for the page to come up the right way round. The rotation instructs the viewer; it never transforms the content.
That has consequences you can measure:
| What survives | Why it matters |
|---|---|
| The text layer | The document stays searchable and the text stays selectable. Ctrl+F still finds what it found before. |
| Hyperlinks and bookmarks | Internal navigation, the table of contents and every external link keep working, because the annotation objects were never rewritten. |
| Form fields | A fillable form is still fillable. Re-rendering the page flattens it into a picture of a form. |
| The file size | Identical to within a few bytes. Nothing was compressed, decompressed or re-encoded. |
| Image quality | Absolutely unchanged, because no image was decoded. A 600 dpi scan is still a 600 dpi scan. |
| Digital signatures | A signature covering the document will be invalidated by any change, including this one. That is the standard working correctly, and it is unavoidable in any tool. |
The alternative, which is what happens when you rotate by printing to PDF or by exporting from an image editor, is to rasterize each page into a bitmap, turn the bitmap, and build a new PDF around it. Everything in the table above is destroyed in one pass. The text becomes pixels, the links become nothing, and a text document that was 300 KB becomes 12 MB of JPEG. It also looks slightly worse, because the page was rendered at whatever resolution the exporter chose rather than at the resolution your reader would have chosen.
People do this by accident constantly, usually because the only rotate button they could find happened to live in a print dialog. If your rotated PDF suddenly got ten times larger and you can no longer select the text, that is what happened to it. Go back to the original. Rotate it here instead.
One detail trips up other tools. A page that is already sideways in the original carries a /Rotate value before you touch it, so this tool adds your turn to whatever is sitting there instead of replacing it, and a page stored at 90 degrees that you turn once more ends up at 180 and looks exactly the way the preview showed it. A tool that overwrites will appear to work on a normal document and silently do the wrong thing on a scan somebody already corrected once.
The realistic version of this job is almost never "rotate every page". It is a hundred page scan in which the pages fed through the document feeder in two orientations, because half the batch was portrait letters and half was landscape spreadsheets, or because somebody reloaded the tray the other way round. Turning all of them fixes the eighty and breaks the twelve.
Stand every page upright in the toolbar above solves exactly this. It measures each page and turns only the ones that are currently wider than they are tall, so a hundred page scan carrying twelve landscape sheets gets twelve rotations and eighty-eight pages left completely alone. One button. On a mixed scan it is usually the only one you need to press.
Three things about it decide whether you can trust it on a document you have not read:
Open one page full size before you run it. Click any tile and the page fills the screen at the resolution your display can actually show, with arrow keys to walk through the document without dropping back to the grid. Ten seconds there is what stops you shipping a hundred page contract that is upside down from page forty.
The usual architecture for a free rotate PDF online service is an upload form, a queue, a command line utility running on a server you do not control, and a temporary download link that expires. Your document crossed the internet, was written to a disk in a building you have never seen, and was kept there for however long a privacy page claims. For a holiday itinerary, fine. For a signed contract, a medical record, a tenancy agreement or a set of payroll scans, that is a data transfer you did not intend to make and probably could not justify if anybody asked you to.
There is no upload endpoint here, because there is no server-side step to upload to. Your file is opened with the browser File API, parsed by a PDF library running as JavaScript on your own processor, and the rotated document comes back as a Blob your browser downloads from itself. The file exists on your disk and in your RAM. Nowhere else.
One consequence of that architecture should be stated plainly, because it is the price of the promise and not a bug: everything happens in one browser tab, so a very large document is limited by the memory on your own machine and never by our servers. A two hundred page text document is nothing. A two thousand page scan at 600 dpi will make a phone struggle, and the honest advice is to do that one on a desktop.
The document is parsed as it lands, so problems appear on the card before you spend an operation rather than halfway through one.
| Symptom | Cause | Fix |
|---|---|---|
| The card says Locked | The document has a password on it. Encrypted content cannot be read without it, and guessing is not a feature we will ever build. | Open it in a reader with the password, save an unencrypted copy, then rotate that. |
| The rotation looks wrong in one reader only | Some older or embedded viewers, particularly inside e-signature platforms and email previewers, ignore the page rotation entry. | Nothing can be done about a reader that ignores the standard. If a specific platform must see it upright, that is the one case where re-rendering the page is the right answer. |
| The preview is upright but the download is not | Almost always the wrong file was opened afterwards. The downloaded name ends in -rotated. | Check the file name in your downloads folder. The original is untouched by design. |
| Pages render as grey plates | The page renderer was blocked by an extension, or failed to load. The rotation itself does not depend on it. | The tool still works. Rotate blind if you have to, but it is worth switching off the blocker so you can see what you are doing. |
| A signature warning appears afterwards | The document carried a digital signature. Any modification invalidates it, which is the entire point of a signature. | Rotate before signing, never after. If it is already signed, ask the signer for a correctly oriented original. |
| The tab runs out of memory | A very long, very high resolution scan. Page previews are held as bitmaps. | Use a desktop rather than a phone, or split the document first and rotate the halves. |
Rotation is rarely the whole job. It is usually the first step in making a scan presentable, and these are the tools people reach for next:
All of them run the same way this one does: your file is read by JavaScript on your own machine, the work happens on your processor, and the result is handed back as a download that never existed anywhere else.
Drop the file into the zone above, turn the pages that are wrong using the arrows on each tile, and press Rotate PDF. The rotation is written into the page dictionary of the document itself, so the file opens upright in every reader for everybody you send it to. It sticks. The rotate button in Chrome, Edge, Safari or Preview does something different: it turns the page on your own screen only, and forgets it the moment you close the tab.
Because your reader gave you a view rotation instead of a document rotation. Most readers offer a rotate control that changes what is on screen and leaves the file untouched, and almost none of them mention that anywhere in the interface you were looking at. Every page in a PDF carries a rotation value in its own dictionary, and only a tool that writes that value changes the document. This one does.
No to both, because nothing is re-rendered. Rotating a page here changes a single integer in the page dictionary and leaves every content stream byte identical. The text stays selectable, the links keep working, images are not decoded and the file size stays within a few bytes of the original. The method that does cost you quality is rotating by printing to PDF, which rasterizes each page into a bitmap and can multiply the file size by ten.
Yes, and that is what the tool is built around. Every page appears as its own tile with its own arrows, so you can turn page 7 without touching page 8. There are also bulk buttons for turning the whole document at once and for standing only the landscape pages upright, which is the usual fix for a mixed scan.
Press Stand every page upright. It reads the dimensions of every page in the document and turns only the ones that are currently wider than they are tall, so portrait pages are left alone. It measures the whole file rather than only the pages whose previews have rendered, which matters on a long document. It cannot detect an upside down page, because an upside down portrait page is still portrait, so open a couple at full size to check.
No. The rotation runs in your browser, on your own processor, and your document is never transmitted anywhere. Confirm it yourself: load this page, disconnect from the internet completely, pull the cable if you have one, and rotate a document anyway. It still works, because there is no server-side step for it to depend on.
No. Rotation does not touch content streams or annotation objects, so the text layer stays searchable, links and bookmarks keep working, and a fillable form is still fillable. The one thing that does break is a digital signature covering the document, because any modification invalidates a signature. That is the signature standard doing exactly its job, and no tool avoids it, so rotate first and sign afterwards.
Not while it is encrypted. A document with a password on it cannot have its pages read without that password, and a tool that could would not be a security tool worth trusting. Open the file in a reader with the password, save an unencrypted copy, and rotate that copy here.
Exactly four: 0, 90, 180 and 270 degrees, always clockwise. The format has no way to store an arbitrary angle as a page property, so nothing can straighten a page that was scanned three degrees off true. That job needs the page content itself redrawn, which is a different operation with a real quality cost, and it is not what this tool does.
No page limit, and no watermark on the result. The practical ceiling is your own memory, since previews are held as bitmaps in the tab: a few hundred pages is comfortable on a laptop and a two thousand page scan at high resolution is not comfortable on a phone. File size is capped at 100 MB per document, and an operation is one run regardless of how many pages went into it.
Yes, and it is a genuinely good way to fix a document somebody just emailed you. Every control is a real button instead of a gesture, the page tiles are sized for a thumb, and the full size preview fills the screen. Keep the documents modest. A phone tab has far less memory headroom than a desktop one, and a long scan at high resolution is what runs it out.
Not for your first operation. Anonymous visitors get one operation per 12 hours counted by network address, a free account gets two, and Premium removes the counter and raises the file ceilings. An operation is one rotation run regardless of how many pages it covered, so turning three hundred pages counts once.
All of these run the same way: your file is read by JavaScript on your own machine, the work happens on your processor, and the result is handed back to you as a download that never existed anywhere else.
Find the apps that slow your store down, by name and in milliseconds, then answer shoppers' doubts with trust badges under Add to cart. Both are free, with no card and no access to your store admin.
Paste any page and get its three Core Web Vitals, the apps costing time, like Klaviyo or Judge.me, and the Shopify or WooCommerce fix.
Show the promises that settle doubts, from free shipping to a money-back guarantee, right under the Add to cart button. One line in your theme.
Change the link after printing and count the scans. 5 tracked codes on a free account, static codes unlimited.
Anonymous visitors get one operation per 12 hours, per network address. That is not a sales tactic, it is the only lever we have: this tool runs on your processor, not ours, so the run itself earns us nothing. An account tells us who you are, and the ceiling goes up the moment you have one. Your allowance resets in 12 h.
No credit card, no trial period, no sales call. Join with Google, GitHub, Facebook or email.
Free accounts get two operations per 12 hours, and yours reset in 12 h. Waiting costs you nothing and is a completely reasonable choice. If PDF work is a weekly job rather than a one-off, the counter is the wrong tool for you and Premium takes it off the page.
Cancel anytime. Everything you have already produced stays yours either way.