Image to PDF — Merge Images into One PDF Online, Processed Locally

How to use

1. Drag images onto the page, or click to pick several at once. You can also paste a screenshot with the paste shortcut, or paste a link into the image URL box if that is all you have. The tool identifies the real format from the file header, so a wrong extension is still detected; animated images contribute only their first frame, and portrait photos taken on a phone are straightened from their orientation tag first.
2. In the image list and sorting panel, drag cards to change the PDF page order, or use the move up, move down, move to top and move to bottom buttons. Each image can be rotated ninety degrees or removed on its own; you can also select all and remove in bulk, or rotate and flip every image at once. The preview on the right immediately reflects the new order.
3. Open PDF paper settings to choose the paper, orientation and margins; with the follow-image-size option, every page may have different dimensions. Then choose contain, cover with cropping or original-size embedding under image fit; if you want stretching that distorts the image, turn on the allow switch first. To place several images on one page, pick a grid under advanced layout; very tall screenshots can use automatic splitting, and page numbers and headers can be set at the same time.
4. Under quality and encoding, decide whether to enable lossless embedding: eligible JPG and PNG files are written through unchanged, while everything else uses the quality slider and encoding format to control file size. Finally, flip through the paginated preview on the right to confirm, enter a file name and click download PDF. If you need to hand the result to code, click get Blob, or send it straight to the browser print dialog.

What is it

A PDF page is essentially a sheet of coordinate paper with a fixed size, measured in points where one point equals one seventy-second of an inch; A4 is 595.28 by 841.89 points. Inside a PDF an image is a separate image object, and the page content stream only records a transformation matrix telling the reader which rectangle to place that image in. Conversion therefore does not rewrite pixels into some other kind of pixels; it chooses a target rectangle for each image and embeds the original image data in the file. How that rectangle is chosen depends on the paper, margins and fit rule: contain takes the smaller of the width and height scale factors so the whole image is visible but leaves white space; cover takes the larger factor and hides the overflow with a clipping path; stretch simply writes the width and height of the usable area, distorting the proportions; original size converts pixels to points at the chosen resolution without scaling up or down. Embedding then splits into two routes. JPEG and PNG are already natively supported image encodings in PDF, so their bytes can be written through unchanged, with no decoding and no second compression, preserving both quality and file size. WebP, BMP, GIF and any image that needs rotation, flipping, resizing or slicing must first be decoded onto a canvas and re-encoded. Understanding the difference between these two layers matters: rearranging pages costs almost no quality, while re-encoding does. One more detail: PDF readers generally ignore EXIF orientation tags, so phone photos carrying orientation information must have that rotation baked into the pixels before embedding, otherwise the image will appear sideways on the page.

Origin & history

PDF was born in 1991 from Adobe's Camelot project, whose goal was called the universal document: open a file on any machine, in any software, with any fonts, and the layout should look exactly the same. Acrobat and PDF 1.0 arrived in 1993. The page description inherited PostScript's imaging model, with coordinates measured in points of one seventy-second of an inch, which is why A4 is 595.28 by 841.89 points and one millimetre is about 2.83 points today. Putting images into a PDF was not easy in the early days: the image had to be encoded as a stream object with DCT or Flate, registered in the page resource dictionary, and finally positioned and sized with a transformation matrix. That structure later became an advantage: an image inside a PDF is an independent object and the page only handles placement, so the same image can be reused across pages and scaled without decoding again. When digital cameras spread in the late 1990s, scanners and photo libraries began packaging batches of pictures into PDFs for archiving, and image-to-PDF became a requirement in its own right for the first time. In 2008 Adobe handed PDF 1.7 to ISO as ISO 32000-1, turning a company specification into a public standard, and browsers gradually built in their own readers. The real turning point came after HTML5: the canvas could read pixels and export JPEG and PNG, Fetch could retrieve remote images, and WebAssembly let parsing libraries run inside a page. Later, libraries such as pdf-lib let front-end code create pages, embed image streams and write clipping paths directly, while pdf.js rendered the generated bytes back to the screen immediately. So layout work that once had to be sent to a server can now close the loop in local memory: originals never leave the device, the preview is the finished product, and download hands you the file. It is worth noting that after more than thirty years the format still uses the same point-based coordinates and converts margins from millimetres, while users still agonise over the same questions from the 1990s: leave white space or not, will it be cropped, is it sharp enough. What has changed is that those choices can finally be seen instantly.

Related keywords