Images to PDF

A4, Letter, or the size of each image

Either way it is a container for your images, not a page layout program.

Pixels in, page size out, when pages match the image

That last row is why the default is a paper size. In matching mode a page is only A4 if the image is 595 × 842 pixels, and at that size print looks soft. In A4 or Letter mode the image keeps all its pixels and is scaled onto the page, so you get the paper size and the print resolution together.

In A4 or Letter mode, wide images get landscape pages and tall ones portrait pages, all the same paper size. In matching mode, different image sizes give different page sizes, and "fit to page" printing then scales each page differently, so identical receipts can come out different sizes on paper. If that matters, make every image the same pixel dimensions first with resize image.

In matching mode the image fills the page edge to edge. In A4 or Letter mode it is scaled to fit and centred, so there is white space on two sides when the image is a different shape from the paper.

When a file is skipped

Any image your browser can open is accepted, and iPhone HEIC photos are decoded on your device first. A file that is not an image is skipped with a short message. Two cases worth knowing:

• TIFF from a scanner or a fax service. Added to the list like any image, but only usable if your browser can open TIFF files, and many cannot. If it cannot, that image is left out and a message names it; the other pages are still built. Export JPG from the scanner software instead.
• A PDF. Understandable mistake. That belongs in merge PDF.

What happens to file size

The two formats are handled differently, and it shows up in the output size.

A JPG is stored exactly as it is. Nothing is re-compressed, so the document is roughly the sum of your files plus a little structure. Every other format, PNG included, is turned into a JPG at quality 92 before it goes in. For photographs that usually makes the document smaller. For screenshots and anything with fine text it softens sharp edges a little, and any transparent area is filled with white before it goes in.

For photographs, cutting the width down with resize image before you build is the only real saving available, because compress PDF rewrites structure and never touches the pictures.

Rotation, and what travels inside your photos

A sideways photo stays sideways. Many phones record the picture the way the sensor saw it and add a separate orientation flag telling viewers to turn it. For JPG files that flag is not applied here, because the JPG goes in unchanged, so a photo your gallery shows upright can land in the PDF on its side. Fix the image rather than the PDF: the metadata tool re-encodes a photo from its raw pixels and has a manual rotate control, so the correct rotation is baked into the file itself.

JPEG metadata comes along for the ride. Because each JPG is stored unchanged, whatever is inside it is now inside your PDF too — the camera and phone model, the exact timestamp, and often the GPS coordinates of where you were standing when you took it. That is fine for your own records and much less fine on a document you are emailing to a landlord or a claims department. Strip it before you build the PDF, again with remove EXIF data.

When not to use this

• You need the text to be selectable. A photo of a page is still a photo. The words cannot be copied, searched or read by a screen reader, and nothing here performs OCR. If the recipient needs to search the document, export a PDF from the original file instead of photographing it.
• You want several images arranged on one page. One image, one page, always.
• The images are enormous and you are on a phone. Every file is read into memory and the whole document is assembled at once. Build it in two halves and join them with merge PDF if a large batch stalls.

Photographed receipts for an expense claim

1. If they came off an iPhone they can go straight in: HEIC photos are converted on your device as the PDF is built.
2. Strip the location data with remove EXIF, and rotate any that are lying down while you are there.
3. Bring them to a sensible width with compress image. A receipt is legible long before it needs 4000 pixels, and this is where the file size is won.
4. Add them here in date order — they are placed in the order you add them — and create the PDF.
5. If finance wants one file and you have several, join them with merge PDF.

Why this one in particular should not be uploaded

Look at what people photograph and turn into PDFs. A passport page for a visa application. A driving licence for a car hire. A bank card for a dispute. A signed form. A child's birth certificate. A prescription. These are exactly the images used to impersonate somebody, and they arrive at a conversion service pre-sorted and labelled.

Here the images are read by your browser, the PDF is assembled by your processor, and the finished file is handed to your downloads folder. There is no upload step and nothing is kept, which also means the result exists only where you saved it — reload the page and it is gone. Every tool in this collection works that way, including the image tools and the rest of the list.