How to Scan Documents to PDF on a Phone, Windows PC or Mac

On an iPhone the scanner hides in two places. In Notes, open a note, tap the attachment button (a camera or paperclip icon, depending on your iOS version) and choose Scan Documents. The camera finds the page edges on its own, flattens the perspective and takes the shot when the page is steady; keep going and every page lands in the same scan. Notes has had this since iOS 11, and since iOS 13 the Files app offers the same scanner from the three-dot menu, which saves the PDF straight into a folder instead of a note.

On Android, Google Drive is the scanner most phones already have. Tap the camera icon or New → Scan, photograph each page, crop and choose a filter, then save; Drive stores the result as a PDF in your Drive. On a computer with a flatbed or all-in-one printer, Windows has the free Windows Scan app from the Microsoft Store, and every Mac includes Image Capture in the Applications folder. Both talk to most scanners without the maker's software.

A flatbed is still the better tool for anything that will be printed, archived or examined closely. The glass holds the page perfectly flat, the light is even across the whole sheet, and the resolution is exact: at 300 dpi an A4 page becomes 2480 × 3508 px and a US Letter page 2550 × 3300 px, every time. Bound books, thick passports and documents with fine print or embossed seals come out cleanly. A sheet feeder on an all-in-one printer also handles a 30-page contract while you do something else.

A phone wins on speed and availability, and for most everyday paperwork the result is just as good. A 12 MP camera, common on recent phones, captures 4032 × 3024 px, which is more than enough detail for a full page of text. What the phone cannot control as well is the light and the angle, so shadows, glare and a slightly bent page are the usual flaws. The scan apps correct perspective, but they cannot invent detail that a shadow hid.

Light matters more than the camera. Put the page on a dark, plain surface such as a desk or a book cover, so the app can find the edges, and stand with a window or a lamp to the side rather than behind you, where your own shadow falls across the paper. Turn the flash off for glossy paper and plastic cards, because it leaves a bright hotspot over exactly the words you need. Hold the phone parallel to the page and fill most of the frame; if the app shows a tilted outline, straighten up instead of relying on the correction.

Choose the color mode on purpose. Grayscale keeps soft pencil marks, faint stamps and signatures readable and is usually far smaller than color. Pure black and white, the mode some apps call B&W or Document, is smallest of all and looks crisp on typed text, but it can wipe out light handwriting, gray backgrounds and photographs on an ID. Keep color only when color carries meaning: a red stamp, a colored form field, a highlighted passage or a photo.

For resolution, 200 dpi is enough for a page that will only be read on screen, and 300 dpi is the safe default for printing, archiving and text recognition. Going higher mostly makes the file bigger. The table below shows the combinations that cover nearly every document, with sizes phrased as typical for one page, since paper and content change the result.

Sometimes the pages already exist as photos: a camera roll of receipts, a few JPGs a colleague sent, or HEIC pictures from an iPhone that skipped the scanner. Most offices want one PDF with all the pages in order, not ten loose images. The image to PDF converter takes up to 50 images of up to 95 MB each, including JPG, PNG and HEIC, and places each one on its own page of a single file.

Two settings make the result look like a scan rather than a photo album. Set Page size to A4 or Letter, so every page has the same paper format even if the photos had different shapes, and leave Image on page on Fit, which scales each photo to the page without cropping. Turn on Auto orient so pages shot sideways come out upright. If a photo shows a lot of desk around the paper, crop it first; the page will fill the sheet and the text will be larger.

If you already scanned some pages to PDF and others are photos, convert the photos first and then combine the files. Our guide on how to merge PDF files covers putting several PDFs into one and moving pages into the right place. The order you settle on now is the order a reader will page through, so it is worth checking before you send.

A scan is a picture of every page, so it is heavy compared with a PDF exported from a word processor. A ten-page color scan at 300 dpi can easily reach 10–20 MB, while Gmail caps attachments at 25 MB per message and many application portals accept only 200 KB, 500 KB or 1 MB per document. The cheapest savings happen before the PDF exists: scan text in grayscale, stay at 200–300 dpi and leave out blank backs of one-sided pages.

When the file is already made, compress the PDF rather than rescanning. Compress by on Recommended recompresses the page images and leaves the layout alone, which is often enough for email. When a portal names a number, choose Target size and enter it in KB or MB, a little below the stated limit. Convert to grayscale is the biggest single saving on color scans of black text, because it throws away color information that never carried anything.

If you are building the PDF from photos, you can do both jobs in one pass: in the image converter, set PDF file size to Target size and the finished file comes out at or under your limit. Very small limits and many pages do not mix well, so for a 100 KB cap send the pages the form actually asks for. For the trade-offs between size and sharpness, read how to compress a PDF without losing quality.

A scanned PDF is a photograph of text: it looks right, but you cannot search it, copy a sentence or let a screen reader read it. Optical character recognition (OCR) fixes that by recognizing the letters and storing them as an invisible text layer behind each page image. The page still looks exactly like the scan, but Ctrl+F or Cmd+F now finds words in it. Some scan apps add this layer automatically and some do not, so test your file by searching for a word you can see on the page.

If the search finds nothing, run the file through OCR PDF, keep Output on Searchable PDF and choose the Document language, which matters for accents and non-Latin scripts. Recognition quality follows scan quality: 300 dpi, an even light and straight lines give clean text, while a shadowed, tilted phone photo produces errors. Enhance contrast helps with faded originals. Do OCR before compressing hard, since a heavily compressed page gives the recognizer less to work with.

A good file name saves the recipient a question. Use a pattern that sorts by itself and says what is inside, such as 2026-08-26_lease_signed.pdf or smith_passport_bio-page.pdf, with the date written year first so files line up in date order in any folder. Avoid names like scan0001.pdf or IMG_4821.pdf, which say nothing, and avoid slashes, colons and emoji, which some portals reject during upload. If a form asks for each document separately, make one PDF per document rather than one PDF of everything.

Check the page order and rotation before you send: open the PDF, scroll from the first page to the last, and confirm that the signature page is at the end, every page is upright and nothing is cut off at the edges. A single missing page is the most common reason an application bounces back. Two-sided documents need both sides in sequence, front then back, for every sheet.

Scans of IDs, bank statements and medical letters contain exactly the details fraud depends on, so think about where they go. Use an online tool that states plainly what happens to your files: on UltKit, uploads are processed and deleted 30 minutes after processing, with no account and nothing to install, as described on how it works. Delete leftover scans from shared computers and downloaded folders once you have sent them, and send the PDF through the portal or a direct message rather than a public link.