PDF to images: get JPG or PNG out of any PDF

Do you need the page, or the picture on it?

Every “PDF to image” job is one of 2 jobs, and telling them apart takes 10 seconds.

Rendering photographs a page. The converter draws everything on it, text, lines, tables and pictures, into 1 flat image at a resolution you pick. That is what PDF to JPG and PDF to PNG do. Render when you want the page as a reader sees it: a page for a slide, a cover thumbnail for a listing, a flyer posted to chat.

Extracting opens the file and copies out the pictures stored inside it. A PDF keeps each photo as its own object, often at a much higher resolution than the size it is displayed at, and Extract images pulls those objects out exactly as they were saved. Extract when you want the photo itself: the product shot in a brochure, the pictures in a property listing, the scans inside a scanned document.

The 10 second test: say out loud what you want. If the sentence is “page 3 as a picture”, render. If it is “the photo on page 3”, extract. If the document is a scan, the 2 answers meet in the middle, because each page of a scan is 1 big stored picture; extraction then returns exactly what the scanner saved, and rendering returns the page as the viewer shows it, including any rotation or cropping the scan app applied on top.

The difference also decides your quality ceiling. A rendered page can never be sharper than the DPI you chose when you rendered it; the pixels you did not ask for do not exist. An extracted original is exactly what the author embedded; for JPEGs it comes out byte for byte identical. So “PDF to image without losing quality” has a precise answer: extract when you want the photo, render at 300 dpi when you want the page.

Render pages as JPG or PNG

The whole job is 4 steps:

1. Drop the file. It shows in the file list with its page count; nothing has been sent anywhere.
2. Pick the pages. Leave Pages on “all”, or type the ones you need, such as 2, 5-7.
3. Set the resolution. 72, 150, 300 or 600 dpi (150 is the default); the next sections say which. For JPG you can also set the quality (85 by default); for PNG, a white or a transparent background.
4. Make the images and download. 1 page downloads as the image itself; several pages come down as 1 zip, or one by one with Several pages set to “as separate files”. Files are named by page number, padded (report-07.jpg), so page 7 stays page 7 on your disk.

Because the conversion runs on your device, a big file costs you no upload time and no mobile data, and the speed depends on your machine, not on a server queue.

The format choice comes down to what is on the page:

• JPG suits photos and scans. It compresses photographic detail well, so a photo page becomes a usefully small file, and every mail client, chat app and website accepts it.
• PNG suits text, line art, diagrams and logos. It is lossless, so hard edges stay clean instead of growing the faint ringing JPG puts around sharp contrast. It is also the only one of the 2 that can hold transparency: with the transparent background option, whatever the page does not paint stays clear, which is what you want for a stamp or a logo that will sit on a coloured slide.

PDF to PNGTransparent background option. Free, runs on your device.Open

A rule of thumb that survives contact with real files: photo pages go JPG, everything else goes PNG. The full comparison, with the same pages converted both ways and the file sizes side by side, is its own guide in this cluster, and so is the transparent PNG case for logos, stamps and signatures.

Pick the DPI before you convert

DPI (dots per inch) decides how many pixels a page becomes. A PDF page has a physical size, A4 is 21 x 29.7 cm, and the DPI multiplies that size into pixels. The tool offers 72, 150, 300 and 600; the 3 you will use:

DPI	An A4 page becomes	Use it for
72	595 x 842 px	thumbnails, previews, chat
150	1240 x 1754 px	documents read on screens, slides, email
300	2480 x 3508 px	print, and zooming in on detail

Two things follow from the arithmetic. First, doubling the DPI quadruples the pixel count, and the file grows with it, so converting everything at 300 dpi “to be safe” produces images 16 times larger than the 72 dpi version, most of which nobody will ever zoom into. Second, DPI only buys sharpness the source can deliver: rendering a rough scan at 300 dpi gives you a bigger soft image, not a sharper one.

Match the number to the destination. Jane Cooper pastes a chart page from the 48-page report into her slides for a meeting in Vienna: 150 dpi is right, the projector cannot show more. The same report page going to a print shop wants 300 dpi. A cover thumbnail for a document library is happy at 72 dpi. When in doubt between 2 stops, take the higher one; you can always scale an image down, but scaling up invents nothing.

You can also work backwards from the space the image will fill. A card 300 px wide on a website needs roughly 600 px of image to look sharp on a high density screen, and 72 dpi already delivers that from an A4 page. A page that will fill a slide wants the 1240 px that 150 dpi gives. Everything beyond what the destination can show is file size spent on pixels nobody sees.

Extract the original pictures instead

When the thing you want is a photo that lives inside the PDF, do not photograph the page it sits on. The page shows the photo at its layout size, say 800 px wide in a brochure column, but the file usually stores the full original, 3000 px or more. Rendering gets you the 800 px appearance; extraction gets you every pixel the author embedded.

Extract imagesEvery picture at its own resolution, JPEGs as they are. Free, runs on your device.Open

The extractor walks the file and pulls out the stored picture objects. What you get:

• JPEGs come out byte for byte as they were saved into the PDF. No re-encoding, no generation loss, the exact file.
• Other stored formats come out as lossless PNG, so nothing is lost in the handover.
• Repeats come out once. A logo placed on all 48 pages of the report is stored once and extracted once, not 48 times.
• Tiny images are skipped, so you are not buried under bullet dots and border graphics.

This is also the right tool for a scanned PDF. A scan is pictures already, 1 big image per page with a thin PDF wrapper around it, so extraction returns exactly what the scanner saved, at the scan’s own resolution. Karol Michał scans 12 receipts in Łódź into 1 PDF and gets 12 images back out, in order, ready for the expense tool.

One case flips the advice: vector content. A chart or logo drawn as lines and fills is not a stored picture, it is drawing instructions, so there is no original image to extract. For those, render the page (crop to the object first if you only need the object), and read the honest part below about what rendering does to vectors.

Convert only the pages you need

A 48-page report with 1 useful chart on page 31 does not need 48 images. Type 31 in Pages, or a list like 2, 5-7, and the converter touches only those. The output zip names files by page number, so what you converted stays findable.

Large files are not a problem, they are just slower. The converter renders pages one at a time and lets each one go before starting the next, so a 500-page file needs the memory for 1 page, not 500. Your device’s memory is the only limit there is, and when a file genuinely pushes it, the page tells you so in plain numbers instead of failing silently.

Two neighbouring jobs are worth telling apart from this one. If you want certain pages as a smaller PDF, not as images, that is extract pages, and the result keeps selectable text. And if you are heading the opposite way, turning photos or screenshots into a PDF, start at the pictures to PDF guide.

The honest part: what an image loses

A page turned into an image stops being a document. Before you convert, know what you are giving up:

• The text stops being text. You cannot select it, search it, or copy it out of a JPG. A screen reader sees nothing unless you write alt text by hand. If what you actually need is the content in editable form, that is a different conversion: see PDF to Word, Excel or text, or go straight to PDF to Word.
• Links die. A clickable table of contents or a link to a booking page becomes ink.
• Vector art stops being vector. A logo or chart drawn as vectors is sharp at every zoom in the PDF; rendered to JPG or PNG it is fixed at the DPI you chose, and JPG adds its compression on top. For print-grade reuse of vector art, the right output is SVG, and a PDF to SVG tool is coming to getPDF; until then, render at 300 dpi and keep the PDF.
• JPG cannot hold transparency. Anything that needs a clear background must be PNG.

None of this argues against converting; it argues for converting deliberately. A flyer going to chat loses nothing that chat could have shown anyway. A signed lease turned into JPGs to “send it quickly” loses its searchable text and its standing as the file everyone refers back to, and whoever receives it will ask for the PDF within a week.

So treat the image as a copy for a purpose, never as the document. Send a contract as the PDF, archive the report as the PDF, and convert a copy when a slide, a chat or a listing needs a picture. The PDF remains the master that can produce every future image; the image can never give you the PDF back.

Nothing is uploaded, and you can check

The converters on this page run entirely in your browser; the rendering engine is loaded as WebAssembly and does its work on your device. No account, no file-size plan, no watermark, and your PDF never travels anywhere. That claim is testable in a minute: convert a file once (the first file fetches the engine, 4.6 MB, and your browser keeps it), turn Wi-Fi off, and convert another. It still works, because there is nothing to send and nowhere to send it. That is the whole privacy story, and it is why a confidential lease or an invoice with bank details is fine to convert here: the only machine that ever sees the file is yours.