The transparency question, answered first
PNG is the format people reach for when an image needs a transparent background: logos, icons, diagrams exported from design tools, screenshots with rounded corners. So the first thing that matters in a PNG to PDF converter is what happens to that transparency, and the common answer is bad. Flatten it onto black and a dark logo disappears. Flatten it onto white without saying so and a design that was meant to sit on a coloured page arrives wrong.
Here the image is embedded as a PNG inside the document, alpha channel intact. The transparent areas show whatever the PDF viewer paints behind the page, which is white in every ordinary reader, and stay transparent if the page is placed into something else later.
Several images, one document, in your order
Most people arrive with more than one file. A set of exported slides, a diagram per page, a sequence of screenshots for a bug report or a manual. Each image becomes one page, and the list order is the page order, changed with the arrows beside each row before you build.
That is worth doing deliberately, because it is the part no other step can fix. A twelve-page document assembled in the wrong order has to be rebuilt, while choosing the wrong margin costs one more click.
Size, and when to use JPG instead
PNG is lossless, and a PNG stays lossless inside the PDF. For diagrams, screenshots and line art that is exactly right: text edges stay sharp and flat colour areas stay flat. For photographs it is expensive, since a lossless photograph is roughly ten times the size of a good JPEG at a difference nobody sees. If your images are photographs, convert them to JPG first and then build the document.
Everything runs in your browser. Nothing is sent to a server, there is no watermark, and no account is needed.