Tukuk-OS🖼️ Images to PDF
jpeg · png · webp · gif · bmp → one document · A4, Letter or fit-to-image
Put a stack of photos, scans or screenshots into one PDF document — in the order you picked them, on A4, US Letter, or a page sized exactly to each image — without uploading anything. The PDF is assembled in this tab by your own browser. Close it and no copy of your images exists anywhere but your device.
What “build a PDF from images” actually means
A PDF page is a coordinate system with instructions painted onto it; one of those instructions is “draw this embedded image object across this rectangle”. JPEG and PNG files already are compressed image streams of exactly the two kinds a PDF knows natively (DCTDecode and FlateDecode), so those go into the document byte-for-byte — no re-encoding, no quality loss, no generational damage, whatever a site that “optimises” your images on the way through might do. WebP, GIF, BMP and TIFF are not native PDF image formats, so they are decoded once by your browser's own image engine, drawn onto a canvas, and re-embedded as PNG — lossless in pixel terms (the source's own lossiness, if it had any, is carried along untouched), just not byte-for-byte identical to the original file. The result list under the download button tells you which of the two paths each image took, instead of leaving you to guess.
Page size, orientation and margins
A4 is 210 × 297 mm — the ISO 216 paper size used everywhere except the Americas. US Letter is 8.5 × 11 inches. Both are placed in the PDF as points (1 point = 1/72 inch), the unit every PDF viewer works in, so the numbers are exact rather than “close enough when printed”. “One page per image, sized to it” makes each page exactly as big as that image's pixel dimensions (1 px = 1 pt), which is what you want for a single photo, a scanned document to be faxed or uploaded to a portal that checks the file is “a PDF”, or an image you never want resampled at all. Orientation set to auto rotates each page individually to match its own image — a landscape photo does not force the whole document sideways. The image is always scaled to fit inside the printable area, centred, never stretched and never cropped: aspect ratio is preserved even when width and height disagree with the page.
Honest limits
There is no text layer. A PDF made of images is, to a search engine or a screen reader, exactly what it looks like: a picture. Extracting the text a human can see in it (OCR) is a completely different, heavyweight technology and is not offered here — a scanner app or desktop suite that does OCR is the tool for that job. Very large stacks of images are limited by your own browser's memory, not by a page-count quota: several hundred full-resolution photos can exhaust a tab before the download button appears, and if that happens, split the batch — no file is lost, since nothing left your machine. Progressive JPEGs and 16-bit or CMYK PNG/JPEG files are rejected by the direct embed path and re-encoded via canvas automatically, exactly as the per-file report below the button says. No password protection, no digital signature — those are separate, deliberately hard things to do correctly, and half-measures here would give false confidence rather than security.