What counts as a scanned PDF?

Any PDF whose text is actually an image — produced by a scanner, fax, phone camera, or 'Print to PDF' from an image. Text cannot be selected in the PDF.

How accurate is your OCR on scans?

We benchmark at 95–97% character accuracy on good scans and 88–92% on phone photos. Table structure is preserved in both cases.

What languages are supported?

English, Spanish, French, German, Portuguese, Italian, Arabic, Dutch, Polish, and more. Language is auto-detected per page.

Can I rotate or deskew automatically?

Yes. Pages are auto-rotated and deskewed before OCR. You don't have to pre-process anything.

Scanned PDF to Excel — OCR-Powered, Table-Aware

Turn photographed, scanned, or faxed PDFs into clean Excel sheets. Our OCR reads messy real-world documents — not just clean, digital ones — and preserves table structure.

📸 Drop your scanned PDF or photo

Up to 10 pages free · Auto-deskew + rotation · 10+ languages

Built for real scans, not textbook PDFs

Auto-deskew & rotate

Crooked scans? Upside down phone photos? Albills straightens pages before OCR.

Low-resolution friendly

Works down to 150 DPI — fax-quality scans and compressed email attachments.

Multi-language OCR

10+ languages auto-detected per page, including Arabic, German, French, Spanish, Portuguese.

Table structure preserved

Rows stay rows, columns stay columns — even when gridlines are missing or faint.

Handwriting (beta)

Printed handwriting on forms is supported in beta. Confidence scored per field.

Batch mode (Pro)

Drop a folder of scans. Get one Excel per PDF — perfect for archive digitization.

5 tips for the best OCR results

Scan at 300 DPI when possible. 200 DPI works, 150 is the floor. Higher DPI = more accurate character recognition.
Use "Black & White" or "Grayscale" scan mode. Color scans pick up background paper textures that confuse OCR.
Make sure the page is flat. Book spines, bent paper, and curled pages distort columns.
Good lighting for phone photos. Shadow-free, even lighting; avoid flash on glossy paper.
Let Albills auto-rotate. Don't pre-rotate — our deskew handles it better than manual rotation.

Frequently asked questions

What counts as a scanned PDF?: Any PDF whose text is actually an image — produced by a scanner, fax, phone camera, or "Print to PDF" from an image. If you can't highlight text with your cursor, it's a scanned PDF.
How accurate is your OCR on scans?: We benchmark at 95–97% character accuracy on good scans and 88–92% on phone photos. Table structure is preserved in both cases. See our public benchmark.
What languages are supported?: English, Spanish, French, German, Portuguese, Italian, Arabic, Dutch, Polish, and more. Language is auto-detected per page.
Can I rotate or deskew automatically?: Yes. Pages are auto-rotated and deskewed before OCR. You don't have to pre-process anything.
Can I OCR a very old or faint document?: Yes, with caveats. We apply contrast enhancement before OCR. For extremely faint originals, consider scanning at 600 DPI for better results.