Images to PDF: Practical Tips and Common Mistakes
Reviewed by the FreeOnline.fyi team · Updated 2026-09-14
What actually happens when you press Create PDF
We built this converter for a boring but recurring job: you have six photos of a receipt, three screenshots of an error message and a scanned ID, and someone wants a single PDF. Drop the files in, check the order in the thumbnail strip, pick a page setup, press Create PDF, and the browser hands you one file with one image per page. The Images to PDF Converter does that decoding and rendering locally on your device, so nothing is uploaded and it keeps working if your connection drops. The output is an ordinary PDF, a format stable enough that it opens almost anywhere and is standardised as ISO 32000.
Two limits are worth knowing before you start. You can add up to 20 images, at a maximum of 15 MB each, in PNG, JPEG, WebP, GIF or BMP. Animated GIFs only contribute their first frame, and HEIC files straight off an iPhone camera roll are not on that list — the picker will simply skip them. On iOS, either set the camera to "Most Compatible" so new shots save as JPEG, or export the photo as JPEG before importing.
When it finishes you get a summary line — something like "7 pages · 3.4 MB PDF" — a Preview link that opens the file in a new tab, and the download itself, named combined-images.pdf or after your first image. Nothing is stored, so if you close the tab you start over from your originals.
Page order comes from the thumbnail strip, not the file names
The mistake we see most often: someone selects a whole folder of files, assumes the tool sorts them alphabetically, and ends up with page 3 sitting before page 1. Page order is exactly the order in the thumbnail strip, which is the order in which the files were added or selected. File names play no part in it.
If something lands in the wrong place, use the up and down arrows on that thumbnail — there's no need to clear everything and start again. The × control removes a single image. For a two-sided receipt, shoot the front then the back so they arrive as a pair, then eyeball the strip before generating. On desktop you can Ctrl/Cmd-click in the file dialog to select several at once, but depending on your operating system the dialog may hand them over in the order it lists them, not the order you clicked — so always do a final read of the strip rather than trusting your selection.
On mobile, the file picker gives direct access to your photo library, and pasting an image into the drop zone works too. Dragging a photo straight from a file manager is the fastest route on desktop, but check the thumbnails anyway; a stray screenshot you meant to drop elsewhere is easy to miss until it's page 8.
Choosing a page setup: fit, A4 Portrait, A4 Landscape or Letter
"fit" makes each page the same size as the image itself, so nothing is cropped and nothing is padded. It's the right choice for screenshots and for digital-only submissions where paper size is irrelevant. Its trade-off is that pages can differ in size within one file, which some print workflows and print dialogs handle clumsily.
The other three options — A4 Portrait, A4 Landscape and US Letter Portrait — force one uniform paper size and scale each image to fit inside it, centred, with the original aspect ratio preserved. A tall phone photo on A4 Portrait will leave white bands on the left and right. A wide scan on A4 Portrait will leave bands top and bottom and make the content smaller than it needs to be. If the content is wide, A4 Landscape usually gives you noticeably more readable detail on the page.
Only one setup applies to the whole batch, so don't mix a landscape ID scan with portrait receipts and hope for the best. Rotate or crop the odd one out first, or produce two PDFs. And match the paper to the destination: Letter for US portals and forms, A4 pretty much everywhere else — the receiving office usually states which it wants.
File size is where most submissions fail
An image-only PDF is mostly your original images, so expect the output to be roughly the sum of what you put in, plus a little packing overhead. Twenty phone photos at 3–4 MB each lands near 70–80 MB, which exceeds the attachment ceiling many mail servers enforce and is far past the limits printed on most government, university and job-portal upload forms (commonly in the 5–10 MB range, though you should check the specific portal).
The cheap fix is to shrink the inputs rather than the PDF. Crop to the document, export screenshots as JPEG instead of PNG, and re-shoot a receipt at a sensible size instead of uploading a 12-megapixel photo for something that needs two megapixels of detail. This tool doesn't re-encode or downscale your images, which is good for fidelity but means whatever you feed in is roughly what you get out.
Two format quirks matter here. Transparent PNGs render onto the white page, so a logo exported with transparency looks clean but won't blend with a coloured background. And a GIF contributes only its first frame, so if you need the animation, a PDF is the wrong container entirely. The universal format, and how it stores images, is documented at PDF on Wikipedia if you want the background.
A five-point check before you hit send
First, compare the page count in the result summary with the number of images you added. A mismatch usually means a duplicate or a file the picker quietly ignored. Second, open the Preview and scroll the entire document — order and rotation errors are obvious there and completely invisible in a file name.
Third, zoom in on the smallest text. A screenshot of a full-page letter captured at 1024 pixels wide may be unreadable on paper; recapture it zoomed in or crop to the paragraph that matters. Fourth, check the file size against the destination's stated limit. Fifth, confirm the paper size matches what was asked for — a Letter-only portal will often reject or badly reformat A4.
Finally, keep your originals. A PDF assembled from images is a container, not a backup, and it carries no OCR text layer. If a number on page 4 turns out to be wrong, you'll want the source photo to fix it.
Where this fits, and where it doesn't
Receipt bundles for an expense claim, ID plus proof of address for a form, a week of homework photos for a teacher, annotated screenshots for a bug report, design mockups for a client — anything where the recipient wants one file rather than twelve. If you do this often, it's worth keeping alongside the other free online tools we maintain; when you're assembling mockups specifically, a CSS gradient generator is a handy companion for backgrounds you don't want to license.
What this tool is not: an OCR tool. Every page is an image, so text inside it is neither selectable nor searchable. If a portal needs searchable text or a small file, run OCR first and then combine the pages, or type the document properly.
And for anything that genuinely matters — visa paperwork, tax filings, signed contracts — treat this as a packaging step, not a verification step. Open the finished PDF in your usual reader, check every page in order, confirm the numbers against the originals, and keep those originals on file. We're confident in the ordering and page-setup logic because we use it ourselves, but the final check is yours to make.