What this converter does
A PDF stores a page as instructions: draw this line, place this letter here, fill this shape. A PNG stores a grid of pixels. Converting from one to the other means drawing the page and then saving the picture, and that is exactly what happens here - inside your browser, using the same rendering engine that shows PDFs in browser windows every day.
You get one PNG for every page you asked for, at the resolution you chose. PNG is a lossless format, so the file you download holds precisely the pixels that were drawn: no softening, no blurred edges around text, no colour shifts. That makes it the right choice for pages with text, tables, line drawings or screenshots.
Choosing the right resolution
Resolution is measured in dots per inch, and it decides how many pixels each page becomes. At 72 dpi an A4 page is about 595 by 842 pixels - fine for a thumbnail and nothing else. At 150 dpi it is roughly 1240 by 1754, which is comfortable on any screen and in any document. At 300 dpi it is about 2480 by 3508, which is what a printer wants.
Higher is not automatically better. Doubling the dpi makes the image four times as many pixels and four times as slow to produce, and on a phone with 2 GB of memory a very large page at 300 dpi can be the difference between working and stalling. The tool caps the pixel count per page for that reason and tells you when it has done so. For most purposes 150 dpi is the sensible default, and 300 dpi is worth it only when the PNG will be printed.
Transparent backgrounds
PDF pages usually have no background of their own: the white you see is the paper, not part of the page. If you tick the transparency option, the areas with nothing drawn on them stay transparent in the PNG, which is useful when the page is a logo, a stamp, a signature or a diagram you want to lay over something else.
Leave it unticked and the page is drawn on white, which is what you want for ordinary document pages. A transparent PNG opened in an app with a dark background will show the text in its original colour on dark, which is often not what people expect.
One more thing worth knowing: transparency is stored per pixel, so a scanned page that looks white is not transparent at all - the white is part of the scan. The option only helps on pages that were created digitally and genuinely leave the background empty.
Nothing is uploaded, and why that matters
Files are not uploaded - everything happens in your browser. There is no server step in this tool: the PDF is read by JavaScript running in this tab, the pages are drawn on a canvas, and the PNGs are created in memory and handed to your downloads folder. Close the tab and nothing remains.
That is not only a privacy promise, it is also why the tool works with no sign-up and no queue, and why it keeps working on a slow connection. The only thing downloaded is the rendering code itself, once, and only after you choose a file.
When another tool is a better fit
If you want smaller files for a website, PDF to WebP produces the same pages at roughly half the size. If you need photographs rather than pages of text, PDF to JPG is usually smaller again, at the cost of some sharpness around lettering.
If what you actually want is the text rather than a picture of it, PDF to TXT or PDF to Word will give you something you can edit. And if you are going the other way - pictures that need to become a PDF - PNG to PDF does that in the same browser-only way.
For a handful of pages out of a long document, Extract Pages from PDF keeps them as a PDF instead of turning them into pictures, which is usually what a form or an office actually asks for. And if the file is simply too heavy to email, PDF Compressor reduces it without changing the format at all.