Convert PDF to Word
Analyzes the font size and spacing in a PDF to infer headings, paragraphs, and lists, then produces a real, fully editable Word (.docx) file — genuine text, not an image. Works best on clearly formatted documents and relies on the file's actual text layer (not OCR), entirely in your browser with nothing ever uploaded to a server.
Your files are processed on your device and are never uploaded to our servers.
Headings, paragraphs, and lists are inferred from font size and spacing in the file, not real document structure — the result is fully editable in Word, but bulleted and numbered lists are written as plain text with a marker prefix (not true auto-numbering), and it works best on clearly formatted documents.
Drag PDF files here or choose them
You can select more than one file
What this tool does
Rebuilds a PDF's content as an editable Word .docx. Text positions and font sizes on the page are read, headings, paragraphs and lists are inferred from them, and real Word paragraphs are built from the result.
When to use it
- You need to edit a document whose original file you do not have.
- You want to reuse sections of a report without retyping them.
- You are moving an old document into an editable format before updating it.
How to use it
- Put your PDFs on the drop area — several at once is supported.
- Wait for processing; you can cancel any file or retry one that failed.
- Click "Download" per file, or "Download all as ZIP".
A practical example
Input
A 6-page report with main headings, subheadings, body paragraphs and a bulleted list.
Output
A .docx whose text is genuinely editable, with headings as separately styled paragraphs rather than plain text in a large font.
The output is real editable text, not a picture of the page — the substantive difference from a "conversion" that just drops a page image into a Word file.
Limitations
- The structure inference is inherently approximate: a PDF has no concept of a heading or a paragraph, only positioned glyphs. Headings and lists are inferred from font size and vertical spacing, the same way a person skimming the page would.
- Results are best on documents with a clear, consistent visual hierarchy and degrade gracefully on anything more exotic.
- The output is a clean, valid but minimal Word document, not a facsimile: complex tables, multi-column layouts and image placement are not recovered.
- Scanned documents contain no text at all, and the tool warns you about that rather than producing an empty file.
Privacy and processing
All files are processed on your device and are never uploaded to our servers, including the batch run and the bundling of results.
Technical notes
- The output is a hand-built OOXML package, adding no new dependency to the project.
- The line-grouping logic is shared with the PDF to Markdown tool, so the two infer structure consistently.
- Batch processing runs one file at a time rather than in parallel, which keeps memory bounded and the interface responsive on low-end devices.
Frequently Asked Questions
Yes — the output is real Word text (genuine paragraphs and headings), not an image of the page. You can select, edit, and search the text normally.