Conversion guide
Choose the conversion method from the source
The best document-to-Markdown method depends on whether the source contains structure, selectable text, or only pixels.
Start with the file, not the tool
| Source | Best first method | Main risk |
|---|---|---|
| DOCX | Direct structure parsing | Floating layout and tracked changes |
| Digital PDF | Direct text extraction | Reading order and table reconstruction |
| Scanned PDF | Page rendering plus OCR | Recognition errors and page order |
| Photo or screenshot | Image OCR | Blur, skew, glare, and missing context |
When to use a local tool
Use a local workflow when documents cannot leave your machine, when you need thousands of repeatable conversions, or when custom filters are part of the pipeline. Pandoc is useful for structure-rich formats such as DOCX. OCRmyPDF can add a text layer before another extraction step. Local tools require installation, version control, and your own quality checks.
When an online converter is practical
An online converter is useful for one-off work, mixed file types, and users who do not want command-line setup. Before uploading, check the file lifecycle, result retention, OCR processor, account rules, and known limitations. Avoid any tool that promises perfect layout preservation without showing failure cases.
A verification checklist
- Compare every heading against the source.
- Check the reading order on multi-column pages.
- Recalculate important numbers and inspect table columns.
- Search for missing footnotes, captions, and page breaks.
- Keep the source beside the Markdown when the content has consequences.
Use the matching converter
Digital PDF, DOCX, image OCR, and scanned PDF OCR each have distinct input rules and limitations.