PDF to PowerPoint

Only one route has been checked. On 10 October 2026, PowerPoint 16.0 for Windows opened decks this tool made from test PDFs through its automation interface and saved them as PDF, pages matching their sources, and read a three-page test deck's text back. That route switched PowerPoint's alerts off and opened the files read-only, so it would not have shown a repair prompt. No download has been opened with a double-click, and editing has not been tested, so pdf.xyz does not promise that PowerPoint opens the file without asking to repair it, or that the slides can be edited there. This page makes no claim about other presentation apps.

The layout and the images, yes; the fonts, approximately. Each slide is the shape of the first page, each picture sits where the page shows it, and each line of text is a text box at its place with its size, weight and colour, except on a page whose drawing would take more than 32 pictures to stack in order, which comes as one picture of the page with its text in invisible text boxes over it. Fonts are matched rather than copied: a font that comes with Windows or Office, such as Calibri or Arial, keeps its name; a free look-alike of one of those, such as Liberation Sans or Carlito (common in PDFs made on Linux), becomes the font it copies; and any other becomes Arial, Times New Roman or Courier New by style, so a line can come out a little wider or narrower. Lines, boxes and backgrounds arrive as pictures, not as shapes, and links do not come across.

It is converted, as pictures. A scanned page has no selectable text, only a picture of its words, so its slide carries the page as a picture and those words are not editable text; the note beside the download names the pages it happened to. For text as well, run OCR PDF on the scan first and convert the OCR'd copy: its slides carry the recognised words in invisible text boxes over each picture. PDF to Excel, whose file would hold nothing, and PDF to Word, whose document would have nothing in it to edit, still stop at a PDF with no selectable text.

Each line of the PDF becomes its own text box, placed where that line sits, so a paragraph wrapped over five lines arrives as five boxes and a page set in two columns keeps its columns apart. A symbol drawn in a symbol font, such as an arrow, gets a box of its own so the words after it stay where the page has them.

Not if the PDF asks for a password when you open it. Our server is given no password, so the job stops with a message saying the PDF is password-protected, and the page links you to Unlock PDF. If you know the password, remove it there (Unlock PDF cannot recover a password you lack), then convert the unlocked copy here. A damaged PDF is different. If its structure cannot be read, our server first rebuilds the file the way Repair PDF does and converts the rebuilt copy, and the note beside the download says so. If it cannot be rebuilt into a file our server can read, it is not converted, and the message suggests Repair PDF. A PDF whose structure reads but whose page content is damaged is converted, and on such a page a picture can be left out with no note.

No. Our server reads its pages, builds the .pptx and deletes the uploaded PDF once the job ends. The .pptx can be fetched from our server for 60 minutes. Beyond that the link no longer works, and a daily storage rule erases whatever copy is still stored. No account is involved.

Up to 100 MB per file.

One slide per PDF page, in the same order, each the shape of the first page. A photo or screenshot on the page is a picture of its own on the slide, where the page shows it; every line of selectable text is a text box at its place, in its size, weight and colour, with its font matched rather than copied; and the page's other drawing, such as lines, boxes, backgrounds and stencilled images, comes as pictures. A page whose drawing would take more than 32 such pictures to stack in order becomes one picture of the page instead, with its text in invisible text boxes over it, selectable but not shown. A picture too large for our server, or one it cannot decode, is left out, and the note beside the download names its page, except on a page our server cannot read before it is drawn, as in a damaged PDF, where a picture can be left out with no note. A page with nothing on it becomes an empty slide.

Yes, as pictures. A scanned page has no selectable text, only a picture of its words, so its slide carries the page as a picture and those words are not editable text; the note beside the download names those pages. Run the scan through OCR PDF first to have its words as text as well: the converted slides then carry them in invisible text boxes over each picture.