1. Start with the clearest scan available
Use the original scan rather than a screenshot or a repeatedly compressed copy. Pages should be upright, well lit and sharp enough for small printed text to be read at normal zoom.
Clear, typed, single-column documents produce the most reliable results. Handwriting, faint photocopies, tables, forms and multi-column pages need more review after OCR.
2. Extract the text locally
Choose the scanned PDF in the Scanned PDF to Word tool. It renders each page and runs OCR inside the browser, then prepares a DOCX file without uploading the PDF to a Metyris processing server.
This initial tool supports clear English scans of up to three pages. A larger document should be split into smaller sections before conversion.
3. Review the Word document before using it
Open the DOCX in Word and compare it with the original page by page. Check proper names, dates, totals, addresses, reference numbers and punctuation. These details are the most costly to get wrong.
OCR output keeps editable text and page breaks. Rebuild tables, forms and complex layout in Word when needed rather than assuming their original formatting was preserved.
- Compare the first and last line on every page with the scan.
- Search for likely OCR confusions such as 0 and O, 1 and l, or broken words.
- Keep the original PDF alongside the Word file for reference.
4. Know the limits before relying on the result
This tool is an OCR helper, not a certified document conversion service. It accepts unlocked PDFs up to 20 MB and three pages, and currently reads English text only.
Use the DOCX as an editable working copy. Do not rely on unreviewed OCR output for legal, financial, medical, government or other high-stakes decisions, and keep the original PDF as the source record.
Tools for this workflow
- Scanned PDF to Wordextract editable text from a clear English scan locally
- Split PDFseparate a larger document into manageable sections
- Rotate PDFcorrect page orientation before conversion