pdf tool
JPG to PDF
Photographs of receipts, a signed page, a whiteboard: things that need to arrive as one document rather than six attachments. Each image becomes a page on a board, drawn exactly as it will print, and you turn, reorder and drop pages until the document is right. Only then does the PDF get built.
Runs on your device. Nothing is uploaded.
Loading the tool…
How to use it
- 1
Drop the images
Each one becomes a page card showing the real page, not a thumbnail of the photo. Nothing is uploaded.
- 2
Choose the page shape
Match each image gives every page the picture’s own shape and no white border. One fixed size puts them all on A4 or Letter.
- 3
Arrange the board
Drag a card by its grip, or focus the grip and use the arrow keys. Turn a page a quarter at a time, or take it out entirely.
- 4
Make the PDF
JPEGs go in untouched wherever possible, so the file is about the size of the pictures added together.
Match each image, or use one fixed size
Match each image is the default because it is what people usually want: no white bars, no letterboxing, every page shaped like the picture on it. When a photograph declares a print density in its own header, that density is used, so a scan tagged at 300 DPI comes out at its real physical size. When it declares nothing, the page is sized so its longest edge is 297 mm, the long edge of A4, which keeps a phone snap a sensible size rather than a poster. One fixed size is the right choice when the document has to print on real paper or go into a binder.
Why a photo can land sideways
An iPhone shooting in portrait stores the pixels landscape and adds an EXIF tag saying which way to turn them. Browsers apply that tag when they display the image, so the preview looks right; a PDF library reading the raw header does not, so the picture lands rotated a quarter turn with no warning. This tool reads the tag itself, measures the page from the upright dimensions and redraws any tagged photograph through a canvas so it arrives the way you saw it.
JPEG goes in untouched, PNG does not
A PDF can carry a JPEG exactly as it is, with no decoding and no quality loss, and that is what happens to any upright JPEG here. PNG has no such path. The library decodes it to raw pixels, throws away PNG’s own filters and re-compresses it, plus a full size transparency mask, so a PNG page routinely comes out several times larger than the file you started with. If the picture is a photograph rather than a screenshot, dropping the rebuild quality below 100 stores it as a JPEG and the file gets far smaller.
Reordering that works with a finger
Drag to reorder is the obvious interaction and the one most tools get wrong on a phone, because a touch drag on a page that scrolls just scrolls the page. Here the grip is the only part of a card that captures a drag, so the rest of the board still scrolls normally. The grip is also a real button: focus it and the arrow keys move the page, which is how anyone using a keyboard or a screen reader gets the same job done.
Limits
- HEIC, the format an iPhone uses by default, cannot be decoded by any browser. Switch the phone to Most Compatible under Settings, Camera, Formats, or convert the files first.
- A PNG is repacked by the PDF format rather than copied across, so a PNG page usually ends up two to five times the size of the source file. Drop the rebuild quality below 100 to store it as a JPEG instead.
- A photograph that carries an EXIF turn, or one you rotate here, has to be redrawn through a canvas, which re-encodes it. Only an untouched, upright JPEG goes in byte for byte.
- A CMYK JPEG is redrawn as well, because a browser canvas can only produce RGB and copying one across under the wrong colour space inverts it.
- Very large photographs use a lot of memory while the PDF is built, because each one is decoded at full size in turn.
What it does not do
- It does not read the text in a photograph. There is no OCR here, so the pages are pictures.
- It does not straighten, crop or clean up a photo of a document. Crop it first if the desk is in the shot.
- It does not go the other way. PDF to JPG does that job.
Questions people ask
Are my photos uploaded?
No. They are read into memory in this tab, drawn onto a canvas for the previews and assembled with pdf-lib, which runs in your browser. The page makes no network request once it has loaded. Photographs of a signed page or an ID are exactly the sort of thing that should not go to a stranger’s server.
Why is my PDF so large?
Because the pictures are. In the default lossless setting the PDF is roughly the size of the images added together, since JPEGs are copied straight in. Drop the rebuild quality under Advanced options to re-encode them, or run the images through an image compressor first if you want to keep control of the quality.
My iPhone photos will not open.
They are HEIC, which no browser can decode. On the phone open Settings, then Camera, then Formats, and choose Most Compatible; every photo taken after that is a JPEG. For pictures you already have, share them to yourself through an app that converts on the way out, or export them as JPEG from Photos.
Can I put several photos on one page?
Yes. Switch the page shape to One fixed size, then open Advanced options and set images on each page to two, four, six or nine. They are laid out on a grid with the margin you choose, each one fitted inside its cell without being stretched. It is the quickest way to turn a pile of receipts into two pages.
The order is wrong. How do I fix it?
Drag a card by the grip at the bottom left, on a phone as well as a mouse. If you would rather not drag, click the grip once and use the arrow keys: each press moves that page one place and the change is announced for screen readers. Remove a page entirely with the cross.
Why did my photo come out sideways in other tools?
Phones store portrait photos as landscape pixels plus a tag saying to turn them. Browsers honour the tag, most PDF writers do not, so the picture lands rotated. This tool reads the tag itself and redraws those photographs upright, which is why it says a rebuilt image is not copied across byte for byte.
What page size do I get with Match each image?
The page takes the picture’s shape exactly. If the file declares a print density, that is used, so a 300 DPI scan comes out at its true physical size. Otherwise the longest edge is set to 297 mm, the long edge of A4. Every card shows the resulting size in millimetres before you build anything.
Rebuild this as a real document
Take what you made here into documents.
Create a document with OneCraftRelated tools
- Image compressorCompress to WebP or JPEG and compare the result against the original under a divider you drag. Or ask for a size in KB. Nothing is uploaded.
- Merge PDFStack several PDFs, see the merged document before you build it, then save one file. Bookmarks per file, alternating scans, nothing uploaded.
- PDF page size checkerDrop a PDF and read a print report: every page size named against ISO 216 and ANSI, trim box against media box, bleed in millimetres, rotation and clearance.
Built with pdf-lib (MIT).
Written and checked by the OneCraft team. Last checked .