Online OCR for scanned PDFs
Convert Scanned PDFs to Markdown or EPUB
Recover text, formulas, tables, and reading order from scanned books and papers. Sign in to upload a PDF, then choose Markdown for editing or EPUB for reading.
- Recover formulas, tables, and reading order
- Convert and download in your browser
Drop a PDF or EPUB here
Upload one document, then choose the result you want to create.
PDF or EPUB · one file at a timeGet usable text out of scanned pages
Turn page images into content you can search, copy, and use. Choose an output for editing or reading; formulas, tables, and complex page layouts receive OCR and structure recognition.
Extract text you cannot select or copy
Upload scanned books, research papers, textbooks, or lecture notes. OCR recognizes the words in page images so you can search the result and reuse passages. Review unclear text against the source scan.
Markdown for editing, EPUB for reading
Download Markdown to edit text, collect quotations, and build notes. Choose EPUB to read with adjustable text in your ebook reader. Both recover content and structure rather than reproduce the PDF page for page.
Convert and download online
After uploading, choose your output in the task settings and submit the conversion. Download the finished file in your browser and open it in your own editor or reader. No desktop app is required.
Recover formulas, tables, and reading order
PDF Craft analyzes page structure as well as text, including formulas, tables, and multi-column reading order. Compare complex sections with the original before quoting or reusing them.
Common questions
Find answers about formats, results, and pricing.
EPUB or Markdown: which should I choose?
Choose EPUB for reading a book with adjustable text. Choose Markdown for editing, document processing or preparing text for a knowledge base. Neither should be treated as an exact visual copy of the PDF.
How should I check the result?
Compare headings, paragraph order and chapter boundaries with the PDF. For papers, inspect both columns, formulas, tables and footnotes. Correct recognition errors before reusing the text.
What affects OCR quality?
Blurred scans, skewed pages, small print and complex layouts can make recognition harder. Keep the original PDF so you can verify uncertain text; successful processing does not guarantee error-free output.
How much does conversion cost, and how long does it take?
Consult the pricing page for current charges. Processing time varies with the document and service workload; follow the task status rather than assuming a fixed completion time.
Detailed guide: format comparison, review steps, and references
Choose an output format, then check the parts of your document that matter most. OCR and layout reconstruction can introduce errors.
Guide updated
Which output fits your intended use?
| Format and purpose | What should you check after downloading? |
|---|---|
| EPUB: continuous reading and adjustable text | Open the file in your intended reader, change the text size and check chapter navigation and paragraph order. Original page numbers and page breaks may not fit reflowed content; return to the PDF when citing the source. |
| Markdown: editing and further processing | Check heading levels, lists and image references in the editor or knowledge-base workflow you actually use. Verify numbers and units in formulas and tables before splitting the text into chunks. Readable text alone does not establish correct meaning or structure. |
How should you check a conversion from start to finish?
- Inspect the source first. Select representative pages with body text, tables or formulas, note what must be retained and keep the original PDF for comparison. A clean cover does not establish the scan quality of the whole document.
- Choose an output for reading or editing and submit the task. Follow its actual status in the task list. If it fails, record the error before proceeding rather than repeatedly submitting the same file without knowing the cause.
- After downloading, compare the representative pages, then check for missing or repeated paragraphs and incorrect chapter order. Resolve uncertain characters, values and footnotes against the source before quoting the text or adding it to a knowledge base.
What are these format descriptions based on?
The W3C EPUB 3.3 specification defines EPUB as a distribution and interchange format for digital publications. It describes packaging and reading structure; it does not certify the OCR or translation quality of a conversion tool.
distribution and interchange format for digital publications and documents
Excerpt from W3C
CommonMark explains basic Markdown syntax. Rendering extensions such as tables and formulas also depends on the editor or downstream tool you use.