Convert PDF pages to images

Each PDF page is rendered through a PDF interpreter in your browser — the same rendering pipeline browsers use for built-in PDF viewing — and rasterized at your chosen resolution onto a canvas. The canvas output is then encoded as PNG or JPG and queued for download. Because the document is parsed locally, embedded fonts, vectors, and images all render exactly as they appear in the original, and no page content ever crosses the network.

Choose the page range you need rather than converting an entire document: a custom range like 1-5, 8, 12-14 keeps the batch fast and avoids producing files you will never use. Results download individually, so you can grab just the one page you need for a post or slide.

PNG is lossless — every rendered pixel survives encoding, which makes it the right choice for pages with fine text, diagrams, or flat color where compression artifacts would be visible. JPG is lossy and substantially smaller: at quality 85 it suits photographs, scanned pages, and anything headed for the web or social platforms that re-encode images anyway.

Resolution follows the destination. 72 DPI matches screen display — correct for embedding in web pages and emails. 144 DPI gives sharp output on high-density displays and in presentations. 300 DPI is print resolution: a Letter page at 300 DPI rasterizes to 2550×3300 pixels, which is also the setting to use before OCR, since text recognizers benefit from the extra detail.

Social platforms and messaging apps often reject PDF attachments outright but accept images happily — converting the single relevant page to JPG is the standard workaround. Slides and design tools frequently import images more reliably than PDFs, and some presentation workflows rescale PDFs unpredictably while rasterized pages land exactly where placed.

The reverse also matters: images give up selectability. Text in a converted page can no longer be copied, searched, or read by a screen reader, so keep the original PDF alongside any image exports. Treat the images as a delivery format for places PDFs cannot go, not as a replacement for the document itself.

Raster size is arithmetic, and knowing the formula prevents surprise downloads. A US Letter page at 72 DPI is 612×792 pixels — about 485,000 pixels per page, or 1.9 MB of raw RGBA data before PNG compresses it (a text-heavy page compresses far smaller; a photographic page barely compresses at all). At 150 DPI the same page becomes 1275×1650 — four times the pixels; at 300 DPI, 2550×3300 — sixteen times the pixels of the 72 DPI pass. Doubling DPI quadruples pixel count, and the files follow.

So set DPI from the destination, not by habit. Screens and email: 72 DPI, often JPG at quality 85 for photographic pages. Presentations and retina displays: 144 DPI. Print or OCR: 300 DPI, where text benefits from the detail — and where PNG's lossless encoding matters more, since fine type is exactly what JPG artifacts chew on. If a batch is headed for an upload form with a 2 MB cap, run one page at your chosen settings first, look at the reported size, and adjust before converting the rest — a single test page costs seconds; re-exporting forty costs minutes.

Converted pages multiply quickly — one 60-page PDF becomes sixty files — so name and structure them as you go rather than after. Convert only the range you need (1-5, 8, 12-14), download before the next pass, and adopt a naming pattern that keeps sort order aligned with reading order: page-01.png, page-02.png … zero-padded, so file managers sequence them correctly instead of jumping from page-10 between page-1 and page-2.

From there, the pages feed the usual destinations: slide decks that import PNGs more predictably than PDFs, CMS posts needing an inline screenshot of a contract or form, social carousels built from document pages, image galleries, and OCR pipelines that want 300 DPI raster input. Throughout, the PDF remains the master — pages are exports for contexts that cannot take documents, and when the source changes, re-run the range rather than editing exported images by hand. Split first when the source is large, convert the fragments you need, and the whole chain stays small, private, and repeatable.