PDF to Image (JPG/PNG)

Render PDF Pages to JPG or PNG

Each page is drawn onto a canvas by pdf.js and saved as an image. Text, vectors and photographs all become pixels — which is the point, since the result can be dropped into a slide, a chat message or a web page without anyone needing a PDF reader.

Scale, and what it means in pixels

The scale selector offers 1x, 1.5x, 2x, 3x and 4x. 1x renders at the PDF's native 72 points per inch, so the pixel dimensions match the page's point dimensions exactly.

For a standard A4 page:

1x   595 x 842 px      ~72 DPI    on-screen preview, thumbnails
2x  1190 x 1684 px    ~144 DPI    slides, documentation, web
3x  1785 x 2526 px    ~216 DPI    decent print, zoomable detail
4x  2380 x 3368 px    ~288 DPI    large print, fine print inspection

Pixel count grows with the square of the scale, so 4x is sixteen times the memory and roughly sixteen times the file size of 1x. Most people want 2x. Go to 3x or 4x only when someone will print the result or zoom into small type.

JPG or PNG

Pick JPG for anything containing photographs or scans. It compresses continuous tone efficiently, and a 2x page is typically a few hundred kilobytes.

Pick PNG for pages that are mostly text, line art, charts or screenshots. PNG is lossless, so it will not put compression fuzz around letterforms, and flat-colour diagrams often come out *smaller* than the JPEG equivalent as well as sharper.

A rough rule: if you would be annoyed by soft edges on text, use PNG.

One image or a ZIP

A single-page PDF gives you a single image file, downloaded directly.

A multi-page PDF gives you one image per page, bundled into one ZIP so your browser is not firing off forty separate downloads. Filenames inside the archive carry the page number, so they sort correctly.

If you want only page 6, run the PDF through Extract Pages first and convert the one-page result. That skips both the ZIP and the rendering time for pages you do not need.

When a page is too big to render

Browsers cap how large a single canvas may be. The exact ceiling varies — it is lower on iOS and on older mobile hardware than on a desktop — but somewhere above roughly 16,000 pixels on a side, or a total area in the low hundreds of megapixels, allocation fails.

You can reach that with a big architectural drawing at 4x, or an A0 poster at any scale above 2x. When it happens the tool says which page exceeded the limit and suggests a lower scale rather than handing you a blank or half-drawn image. Dropping one step is almost always enough.

Things worth knowing about the result

The text is gone as text. You cannot search or copy from a JPG. That is inherent to rasterising, not a shortcoming here. Keep the PDF if you need the text.

Transparency does not survive JPG. Pages are rendered on a white background, which is what you want for a document and not what you want if you were hoping to composite a logo. PNG output keeps the page background white too — it is rendering a page, not extracting an object.

Embedded images are not extracted at their original resolution. This renders the page as it appears. If you need the untouched source photograph that was placed into the document, a desktop tool that pulls image XObjects out is the right instrument.

Files up to 100 MB are accepted, and rendering happens in your browser rather than on a server. The reverse conversion is Image to PDF.