How do I convert a PDF to Word, PowerPoint, or Excel?

Word and PowerPoint

Both conversions run through the cloud converter: the PDF is sent to the service and the .docx or .pptx downloads with your document’s name. The dialog takes a page range ("1-3, 7" — blank means everything) and an "Include markups" checkbox that flattens your annotations into the source first, so redlines and notes survive into the Word file as part of the page.

Because it is a server conversion, the file leaves the browser for this one — the honest trade for .docx/.pptx fidelity, which layout-aware converters do far better than copy-paste.

Excel — real cells, not screenshots

Table export is different: it runs in your browser. The extractor clusters each page’s text into rows by vertical position and finds the columns where x-positions align across rows; cells that look numeric arrive as real numbers. Every detected table becomes a sheet in the .xlsx — named after its page — so a schedule across five sheets lands as five worksheets.

Two honest requirements: the page needs a text layer (scanned tables need OCR first), and the detector needs at least two rows sharing two aligned columns — genuinely tabular content. A "Use AI table detection (Gemini)" checkbox sends positioned text to the model instead, which handles messier layouts and falls back to the positional pass if it fails.

What to expect

Conversion quality tracks layout complexity: single-column documents come through cleanly, while multi-column layouts, heavy graphics, and scans take more cleanup. If the Excel export reports "no tables detected", the cause is almost always a missing text layer or a layout that is not really a grid.

Doing it in PDFDepth

Header → Import & Export → Convert to Word / Convert to PowerPoint / Export tables to Excel. Pick the range, decide on markups (or AI detection for tables), and the file downloads. For CAD-bound output rather than Office formats, Convert to CAD produces DXF/SVG linework instead.