Convert PDF to Word Online
A neural network looks at each PDF page and finds the headings, paragraphs, tables and charts. You get a Word document with text you can edit.
Files are processed on your device and never uploaded to a server.
- Choose PDF
- Convert to Word
- Download DOCX
Recent results
The first time, the browser downloads a 70 MB neural network. After that it works offline.
Converting to Word…
The graphics card is not available for the neural network on this device, so the processor does the work. It takes longer, but the result is the same.
Done! PDF converted to Word
What's next?
More things to do with your document
How to use
- Click "Choose PDF" or drop the file onto the page. On a phone, the PDF can be up to 100 MB.
- Click "Convert to Word". The first time, the browser downloads a 70 MB neural network and an engine for it, then keeps both in the browser. While a page is being converted, its thumbnail shows what the network found there.
- Under "Word document", click "Download DOCX". The file gets the same name as the PDF.
What you get in the Word document
- Text comes in as ordinary paragraphs, so lines rewrap by themselves when you edit.
- A table becomes a Word table with the same rows and columns. If it runs across several pages in the PDF, Word gets one table, and the header row repeats on each page.
- Charts and photos go in as pictures, in the same places.
- Headers, footers and footnotes work as in Word too. Page numbers in a header or footer update automatically.
How the neural network reads a page
Each page is read by the PP-DocLayoutV3 neural network (PaddlePaddle, Apache-2.0 license). It marks 25 kinds of blocks (headings, paragraphs, tables, charts, headers, footers, footnotes and more) and their reading order.
The network doesn't recognize letters: the text comes straight from the PDF.