Convert PDF to JPG

Why can’t I save my PDF as a JPEG?

Because the question has no single answer, and software would rather offer nothing than answer it wrongly. Think about what each format actually is. A PDF is a document: an ordered stack of pages, any number of them, each free to have its own dimensions. A JPEG is one picture with exactly one width and one height. Point a "Save as" menu at a thirty-page contract and it has to silently decide whether you meant page one, the page you happen to be looking at, or all thirty stitched into a strip — and every one of those guesses is wrong for somebody.

So most readers duck it. Preview on macOS will export a single page and quietly drop the rest. Acrobat hides the option behind a paid Export feature. Chrome, which opens more PDFs than any other program on earth, has no image export at all. None of that means the conversion is hard; it means the menu could not ask you the question. This page asks it: which pages, at what resolution, at what quality. Then it writes one image per page and tells you what each one weighs.

How to convert PDF to JPG without installing Acrobat

Nothing gets installed and nothing gets signed up for. Hand in the file, and the renderer that already ships with your browser draws the pages — the same open-source engine, pdf.js, that Firefox uses to display PDFs natively. It is served from this domain rather than a CDN, and it is fetched the moment you hand in a file rather than when the page loads, so a visitor who never converts anything downloads none of it.

The practical difference from the desktop route is that there is no queue, no export dialog buried three menus deep, and no file leaving your machine to reach a stranger's server. The practical difference from the other web converters is the same thing said from the other side: they need your document on their hardware to do the work, and this does not.

Quality is a number of bytes, not a number out of a hundred

Every converter shows a quality slider. Almost none of them tell you what moving it does, which makes the control close to useless — 78 is not a size, an appearance or a promise. It is a knob you turn while hoping. And you cannot reason your way to the answer either, because the same setting produces forty kilobytes on a page of plain text and nearly two megabytes on a spread of photographs. The variable that dominates is your page, not the number.

So the numbers on the buttons here are measured rather than modeled. Your own first selected page is drawn once at the resolution you have chosen and then encoded at each setting in turn, and the byte counts you see are what those encodings actually came out to. Change the resolution and they are measured again. The one figure that is an estimate — the running total for a whole multi-page document — says "about", because scaling one measured page across the rest is arithmetic rather than observation, and the two should never be dressed alike.

Resolution is what invents the pixels

A PDF page holds instructions, not a picture: draw this rule, set this word in this face at this size, place this photo in this box. Its dimensions are physical ones, recorded in a unit called the point, of which there are 72 to the inch — but a pixel count exists nowhere in the file until a renderer picks one. That pick is the resolution setting, and it separates an image fit for a web page from one fit for a printer.

A standard 8.5-by-11 sheet lands at 612 by 792 pixels if you ask for 72, at 1275 by 1650 if you ask for 150, and at 2550 by 3300 if you ask for 300. The tool prints that figure for every page before it runs anything, and it accepts a target pixel width directly if that is what you actually know. Browsers also cap how large a single drawing surface can be, so an oversized sheet at a high setting is checked against that ceiling in advance and the highest workable figure is named — rather than handing back an image that came out smaller than you asked without saying why.

There is no transparency in a JPEG, so something goes behind the page

A PDF page is empty where nothing was drawn on it. The white you see in a reader is painted by the reader, not stored in the file. JPEG has no alpha channel whatsoever, so those empty areas cannot be carried across — they have to be composited onto something opaque before encoding, and a converter that forgets to do it produces pages that come out solid black.

White is the default here because that is what nearly everyone means by a page. The picker only appears when the document genuinely contains see-through pixels, which is decided by decoding the rendered page and inspecting its alpha values — the only way the question can honestly be answered, since a drawing surface always writes four channels whether or not the fourth one is doing anything. On a fully opaque scan, the control would have nothing to change, so it is not shown at all.

When PDF to PNG is the better call

If your pages are text, tables, forms or line drawings, use the lossless format. JPEG works in eight-by-eight blocks, and the first thing it sacrifices is exactly what a crisp dark stroke on pale paper consists of: abrupt changes from one pixel to the next. What you get back is ringing — faint gray fringes hugging each letter, serifs going mushy, a soft smear along every ruled line. Harmless in a thumbnail, embarrassing at full size. PDF to PNG does this same job without any of that, at the price of a considerably heavier file.

Photographs and scans invert the advice completely. Continuous tone is exactly what JPEG was invented for, and a scanned page will often be several times smaller with no visible difference at all — which is why the tool checks whether your document even has a text layer and adjusts what it recommends accordingly. Undecided?JPG vs PNG works through the trade-off properly. And if the file is right but simply too heavy,compress it to a target size afterwards.

Thirty pages means thirty images

That is not a quirk of this tool, it is what the two formats require, and it is the whole reason the page grid exists. Every page appears as a thumbnail with its own dimensions and its own checkbox, and you can also type something like1-4, 9 if clicking through a long document sounds tedious. Both controls drive the same selection and stay in step with each other.

Finished output arrives the same way: one file per page, each with its own download link, plus a single archive when there is more than one. A lone page never forces you through a zip, and forty pages never force you through forty clicks. Going the other direction — images that need to become one document — isPNG to PDF, which takes JPEG just as happily as it takes PNG.

What this will not do for you

There is no OCR, and the output carries no text layer. A JPEG is a grid of pixels and the format has nowhere to keep words, so nothing in these images will be selectable, searchable or copyable. No conversion to a picture can preserve that, whoever writes it. When the words are what you were after,extracting the text reads the text objects already sitting inside the file, andhow to copy text from a PDF covers the cases where copying misbehaves.

It also does not edit, reorder, delete or merge anything in the source document, and it runs strictly one way. Encrypted files are handled rather than refused: a password box appears with your file still in place, since being asked for a password is a question and not a failure. A file that is corrupt or not a PDF at all is named and turned away with the reason. One page that refuses to draw is flagged in the grid while every other page is still produced, because a long report should not be discarded over a single broken font reference.

Nothing leaves your browser

Converting here involves no upload endpoint, because this site does not operate one for the job. The bytes your browser read off your own disk are handed to a renderer running in this tab, and the images come back the same way. That is a stronger guarantee than a promise to delete files after an hour, because the file was never anywhere that deletion would apply to. Any password you type is used by that same local renderer and goes nowhere either.

None of which you have to take on faith. Open your developer tools, select the Network tab, and run a conversion. You will see the renderer being fetched, possibly some font data if your document needs it, and after that nothing at all going out with your document inside it.