Convert JPG to PDF
Add images, drag them into the order you want, and download one PDF.
How it works
Most of the time this conversion is not something you wanted; it is something a form demanded. Council and government portals, university submission systems, insurance claim uploads, visa applications and expense tools routinely accept PDF and reject JPG, even when the thing being submitted is obviously a photograph of a receipt or a passport page.
Printers are the other half of the demand: a PDF carries an explicit page size and orientation, so it prints predictably, whereas a JPG is handled differently by every driver and print dialog. The images that end up here are therefore usually identity documents, medical letters, bank statements and signed forms. That is a good reason for the conversion to happen in your own browser rather than on somebody's server.
Fit, A4 or Letter
The page size choice matters more than it looks. Fit gives each image a page cut to its own dimensions, so nothing is letterboxed, nothing is cropped, and the image is not resampled at all: the sensible option for screenshots, comic pages, or anything that will only ever be read on a screen. A4 or Letter puts every image on a standard sheet and centres it. An upload form asking for A4, or a print shop, expects exactly that.
The cost is resolution, and it is worth stating plainly: an 800-pixel-wide image scaled to fill the 210 mm width of A4 prints at roughly 96 DPI and will look visibly soft. Print work generally wants 300 DPI, which on A4 means about 2,480 pixels across, so a modern phone photo has plenty and a screenshot usually does not. Adding a margin makes the printed image smaller and therefore slightly sharper, because the same pixels cover less paper.
The 200-inch page ceiling
One hard limit applies to fit and not to the fixed sizes: a PDF page cannot exceed 200 inches, so a full-page mobile screenshot tall enough to pass that gets a page capped at the limit. Nothing is resampled when that happens: the image goes in whole and only the declared page gets smaller, so its resolution rises rather than falls.
Why JPEG, and what it costs
Every image is re-encoded as JPEG before it goes into the PDF, because JPEG is how a PDF stores a photograph most compactly, and the document structure adds only about 360 bytes per page on top. JPEG spends its bits on smooth gradients and leaves visible ringing around the hard edges of small text, so a photograph survives the trip almost invisibly while a crisp screenshot can come out looking slightly dirty. Fit with quality at 95 or above keeps small text legible, since it neither rescales nor quantises much.
Re-encoding can add bytes
The size of the result moves in both directions, which is worth knowing before you start. Photographs usually come out smaller, often substantially: a quality-95 source was 57 percent smaller at the default quality of 85, and a quality-90 source 33 percent smaller. That reverses once a source has already been compressed hard, because re-encoding cannot recover detail that is gone and spends bytes describing the previous compression's artefacts instead.
Break-even sits near source quality 78 for a detailed photograph; below it the PDF grows, by 17 percent from quality 60 and 22 percent from quality 50. Smooth, low-detail pictures such as a plain sky shrank at every source quality measured. Screenshots are the extreme, because PNG stores flat interface colour almost for free: one of 1,440 by 900 and under 9 KB became a PDF of about 250 KB, and the quality-95 setting that keeps its text readable took it past 350 KB. If a PDF is only one of your options, a screenshot is genuinely better left as a PNG.
Thumbnails beat filename sorting
Page order is your decision, not the tool's, and you set it by dragging the entries in the file list rather than by relying on filenames. This matters because filename sorting is a trap: IMG_2.jpg sorts after IMG_10.jpg in most file managers, and camera rolls, scanner outputs and screenshot folders all number without zero padding. The list shows a thumbnail of each image, so you can confirm you have the right page in the right slot before spending a conversion on it.
Images are decoded strictly one at a time and released before the next is opened, so a batch of fifty phone photos completes here instead of exhausting the tab's memory. The same file added twice is decoded once and stored once rather than doubling the output. The progress bar advances once per image, and each step is a real one.
Questions
Are my images uploaded?
No. Each image is decoded by a WebAssembly codec inside a Web Worker in this tab, re-encoded, and written into a PDF that the browser hands you as a download. Open the Network tab in developer tools before you convert: you will see the code and codec files the page needed to load, and no request carrying an image.
Why does my screenshot look soft in the PDF?
Because it is being stretched to fill a paper-sized page, and a screenshot has fewer pixels than a sheet of paper wants. A 1,200 by 800 screenshot is wider than it is tall, so auto orientation gives it a landscape A4 page and those 1,200 pixels spread across 297 mm: about 103 DPI, where a printer resolves several times finer. The same screenshot forced onto portrait A4 spans 210 mm instead and reaches about 145 DPI, still soft. Choose the fit page size and the page is cut to the image, so nothing is scaled and the result is exactly as sharp as the original.
What order will the pages be in?
The order shown in the file list, top to bottom, which you change by dragging or with the move buttons. Nothing is sorted for you, deliberately: automatic filename sorting puts image 10 before image 2 whenever the numbers are not zero-padded, and that is a mistake you would only find after the PDF was submitted.
How many images can I add, and how large?
There is no fixed count. Each image is capped at 50 MB, which is far above any camera JPEG, and a file over the cap is refused at selection time with its size and the limit rather than failing partway through. The practical ceiling is your device's memory, and because images are processed one at a time that load does not grow with the size of the batch — a hundred small images is easier here than four enormous ones.