ScanToDocx turns a photo of a document into a Word file you can edit. Point your phone at a page, drop the picture on the window, and a .docx appears beside it - with the text as real, editable paragraphs rather than a picture of a page.
Nothing is uploaded. The page is found and the words are read on your own computer, so a contract or a payslip never leaves the machine. Your originals are never touched: every document is written as a new file beside the photo it came from.
Want the page kept as a page instead? ScanToPDF is the sibling app for that - same pipeline, ending in a searchable PDF rather than editable text.
Highlights
- Text you can actually edit - the page comes back as paragraphs in a Word document, ready to correct, reuse and send. Not a scan pasted into a page.
- Finds the page and flattens it - the document is detected as a four-cornered shape wherever it sits in the frame and the perspective is unwarped, so a page shot at an angle over a desk comes out square. That is also what makes it readable enough to recognise.
- Rebuilds the structure, not just the words - lines that run on are joined back into paragraphs, a section heading comes back as a real Word heading you can navigate and restyle, a bulleted line becomes a real list item, and centred or right-aligned lines keep their place.
- A table stays a table - a grid on the page is rebuilt as a real Word table with its rows and columns intact, instead of arriving as one loose paragraph per cell that you have to reassemble by hand.
- Turns the page the right way up - a page photographed sideways is read four ways round and the orientation that makes sense wins. It is the single biggest cause of a document that comes back empty.
- Reads over thirty languages - or detects the language page by page. Naming one also sets the document's proofing language, so Word stops underlining every word of a Vietnamese page as a spelling mistake.
- Combines a batch into one document - twenty photos of one contract become one .docx, each page starting on a new sheet. Or stay twenty separate files, because only you know which it was.
- Whole batches at once - drop twenty photos and they queue up, several at a time, each landing beside its own original as it finishes.
- Straight into an editor - a finished document opens in DocCafe in one click to fix whatever the page got wrong, or in whatever your computer opens a .docx with.
- Entirely offline - no account, no upload, no queue on somebody else's server. Works on a plane.

How it works
Drop photos anywhere on the window, or hand them over with the "Open With" command in your file browser. Each one is read, the page inside it is found and flattened, the lighting is evened out, the page is turned upright, the text is recognised, and the result is written next to the original with a -text suffix. There is no confirm step: dropping is the instruction.
The Settings panel is optional and stays out of the way. It is where you name the suffix, pick the language, and decide whether the page is found and turned upright for you. Every row reports what it found - how many words, how many paragraphs, whether a page was detected at all - so a page that came out thin is obvious before you open it.
Perfect for
- Reusing a printed document - a contract, a policy or a form that only exists on paper, back in a file you can edit and send.
- Students and researchers - photograph the pages you need from a library book and quote from text rather than retyping it.
- Notes and whiteboards - a page of handwritten-in-print notes or a typed handout, turned into something searchable.
- Invoices and receipts - get the figures out as text you can paste into a spreadsheet instead of reading them off a photo.
- Anything confidential - the page is read on your own computer, so nothing sensitive is handed to an online converter.