PDF Reader
View PDF files directly in your browser.
Drop a PDF to open and read
or click to browse
View PDF files directly in your browser.
or click to browse
A PDF page has a natural size measured in points, and it is rarely the size of your screen. When a file opens, the reader measures the visible width of the reader container, subtracts 48 pixels so the page is not jammed against the edges, and divides that by the page's natural width. The result becomes the zoom level.
This is why the same document opens at a different percentage on different devices, and why a wide landscape page or an A3 drawing opens at a smaller percentage than a portrait A4 page. It is also why Fit Width is worth pressing after you rotate a phone: the container width changed, so the old zoom level no longer matches. Pressing it recalculates against the page you are currently on, so a document that mixes portrait and landscape pages can be refitted on each one.
Each zoom step asks PDF.js for a fresh rendering of the page at the new scale rather than scaling up pixels that were already drawn. Text and vector line work stay sharp at any level, which is the practical difference between this and pinching into a screenshot. Scanned pages are the exception: a scan is a photograph of paper, so it carries a fixed number of pixels and enlarging it eventually shows the limits of the original scan.
Zoom moves in steps of 25 percentage points, with a floor of 25% and a ceiling of 500%. The floor exists because a page rendered smaller than that is unreadable on any screen; the ceiling exists because the canvas grows with the scale, and past 500% a large page can consume enough memory to make a phone browser drop the tab. If you need detail beyond 500%, export the page as an image instead and open that — the PDF to Image tool renders at up to 3x, which gives you a file you can enlarge in any image viewer.
Both the up/down and left/right pairs move whole pages. Nothing scrolls within a page, because one page is drawn at a time rather than stacked in a continuous strip. That is a deliberate trade: rendering a single page keeps memory flat no matter how long the document is, so a 900-page manual behaves the same as a 3-page letter.
Pages are drawn to a canvas, and a canvas holds pixels rather than characters. Two honest consequences follow, and it is better to know them before you need them.
You can search the document, but not drag-select its text. The built-in search bar matches the PDF's text layer, counts the matches, jumps to each one, and highlights it — even though your browser's own Ctrl+F only searches the interface around the viewer. Selecting and copying text by dragging is not available, because the page is drawn as pixels rather than characters; when you need the words themselves, PDF to Text extracts them and PDF to Word produces an editable document. If the PDF is a scan with no text layer at all, OCR creates one first.
Encrypted files will not open here. A PDF that requires a password to open has to be decrypted before any viewer can draw it. The Protect PDF tool has an unlock mode that removes the password when you know it; open the unlocked copy in the reader afterwards.
The reader also accepts one file at a time, and checks that the browser reports it as a PDF. If a file that really is a PDF is refused, it usually reached you with the wrong type attached — renaming it so it ends in .pdf and re-selecting it normally settles the matter.
The most common use is the last look before an email goes out: confirming the right pages are present, in the right order, the right way up. Opening a viewer in a tab is faster than waiting for a desktop PDF application to start.
After a merge, a split, or a page reorder, the question is always whether the output is what you intended. The page counter answers it immediately — if a merge of a 12-page and a 9-page file does not show 21 pages, something went wrong and you can catch it before sending.
Mobile browsers frequently hand a PDF to a separate app, or download it and leave you to find it in a file manager. Opening it in the browser keeps it in the tab you are already in, and Fit Width sizes it to the screen you are actually holding.
Before running OCR or compression, it helps to see what you are starting from. Zooming to 200–300% shows whether the scan is sharp enough for text recognition to succeed, whether pages are skewed, and whether anything was cropped off at the edge by the scanner.
On a shared, locked-down, or public machine, installing software is often impossible and usually unwise. A browser tab needs no installation, and because the file is never uploaded, closing the tab is enough to leave nothing behind.
After filling a form with Fill PDF Form, viewing the result shows how the values will actually print, which is not always how they looked in the form editor.
Yes. Type in the search bar and the reader matches the document's text layer across every page, shows a running match count, jumps to each hit, and highlights it on the page — with Prev/Next to step through matches. A scanned PDF with no text layer needs OCR first to make its words searchable; to pull the words out as a file, PDF to Text extracts them and PDF to Word gives an editable document.
You can search the text and jump to any match, but dragging to select and copy is not available in the viewer, because the page is drawn as pixels rather than characters. When you need the words to copy or edit, PDF to Text gives you the text and PDF to Word produces an editable document.
Not directly. Use the unlock mode in Protect PDF to remove the password from a file you know the password for, then open the unlocked copy here.
Yes. The document you are viewing is handed to that tool, so it opens ready to work on instead of asking you to find the file again. It is passed inside your browser — the file still never leaves your device, and it is discarded when you close the tab. A workspace bar at the top of the page shows what is loaded, with a Clear button to drop it at any time.
The rendered canvas grows with the zoom level, and beyond 500% a large page can use enough memory for a mobile browser to discard the tab. For finer detail, export the page with PDF to Image at 3x and enlarge that image instead.
Every zoom step redraws the page at full resolution, and a high-DPI scan is a large image to redraw. Zooming out a step or two makes it noticeably lighter, and devices with more memory handle big scans more comfortably than budget phones.
Yes. The Night button in the toolbar inverts the page to light-on-dark for comfortable reading in low light, and your choice is remembered the next time you open the reader. The interface already follows your system dark-mode preference; Night mode extends that to the document. Turn it off to see photographs, signatures, and coloured charts in their original colours.
The PDF stays on your device. Read aloud hands the text of the current page to the text-to-speech voice installed on your device, and that voice does the speaking. Some voice engines synthesise speech on the device itself; others run in that engine's own cloud. Check the voice you have installed if that matters for a sensitive document, or simply read the page on screen instead.
The whole file is read into memory at once, so the practical limit is the memory available to your browser tab rather than a fixed page count. Because only one page is rendered at a time, page count matters far less than the size of the individual pages — a long text document is lighter than a short document full of high-resolution scans.
Privacy: The PDF never leaves your browser. Rendering uses PDF.js locally. Closing the tab unloads everything from memory.