Copied, not converted
Most JPG to PDF tools decode every photo into pixels and compress it again on the way into the document. That second compression is where the soft edges and the blocky sky come from, and it happens even when nothing about the picture needed to change. A PDF can hold a JPEG stream exactly as a camera wrote it, so this page does that whenever it can: the compressed bytes of the photo are copied into the file, minus the metadata blocks, and the page is told where to draw them.
Turning a sideways phone photo upright does not break this. The quarter turn is written as a transform in the page, not into the pixels, so a portrait shot that the phone stored on its side is still copied untouched. Fill and reframing work the same way, as a clipping path around a picture that is otherwise exactly the file you chose.
Know before you save
The Photos step on this page shows a running count — “11 of 12 photos will be copied into the PDF unchanged” — worked out from each JPEG’s own frame header with the same test the writer applies. If the number is lower than you expected, the line under it says why: grey pages, a resolution ceiling smaller than the photo, or a file type that PDF readers cannot draw from its own bytes.
A worked example you can repeat: press “Try six sample photos” above. The six are drawn in your browser as ordinary baseline JPEGs, and the count reads 6 of 6. Saved, they made a 6-page PDF of 189 KB with all six copied straight from the files. In our own test set, three phone-sized JPEGs and one progressive JPEG gave “3 of 4”: the progressive one was redrawn, and the other three streams in the PDF were byte-identical to the originals with their EXIF blocks removed.
When a copy is the wrong answer
A copy keeps the file size of the original. Twelve phone photos at 4 MB each make a PDF of about 48 MB, which is fine for an archive and useless for an upload box. When the document has to be small, set a resolution ceiling or a size on the Quality step, or start from the size-limited builder or JPG to PDF under 200 KB. Only the pages that dominate the file are redrawn; the rest stay copied. For a mix of JPEGs, PNGs and screenshots, Image to PDF shows each file’s format as it goes in.
One limit worth stating plainly: the pages are pictures. Text in a photographed letter cannot be selected or searched in the PDF this writes, copied or not.
Questions about JPG to PDF
Will my JPGs lose quality?
Not the ones that can be copied. A baseline JPEG that needs no crop, no grey and no shrinking is written into the PDF as the same compressed bytes it already was, so there is no second round of compression at all. The builder tells you how many of your photos qualify before you save, and the receipt afterwards says how many actually were copied.
Are my photos uploaded?
No. The JPEGs are read from your disk into this tab, and the PDF is assembled by code already running in it. Copying a JPEG into the document is a byte operation inside the browser; there is no server on this site that could receive one.
What does the PDF contain?
One image per photo, placed on its page, and nothing else: no text layer, no title or author field, and no EXIF. The camera model, the date and any GPS position are removed from each JPEG as it is copied in — the picture itself is untouched.
Which JPEGs cannot be copied?
Progressive JPEGs, CMYK JPEGs, 12-bit files and photos whose EXIF says they are mirrored. Many PDF readers draw those wrongly from their own bytes, so they are redrawn once at quality 0.82 instead. A resolution ceiling that is lower than what a photo carries also means a redraw, which is why this page opens with no ceiling.
Does a JFIF file work?
Yes. A .jfif file is a JPEG with a different extension, and files are recognised here by their bytes rather than by their names, so it goes in and is copied exactly like any other JPEG.