PDF в Excel

Извлеките таблицы из текстовой раскладки PDF в книгу Excel, один лист на страницу — полностью в браузере.

Это использует позицию каждого слова, чтобы угадать строки и столбцы — это не настоящее распознавание структуры таблицы, поэтому сложные или многоколоночные макеты могут не выровняться идеально. Проверьте предпросмотр перед скачиванием.

PDF обрабатывается в браузере — ничего не загружается, пока вы не скачаете.

Перетащите или нажмите, чтобы выбрать PDF

Best practices

- Works best on simple grids like invoices, price lists, and reports with consistent column alignment.

- Multi-column page layouts (like a two-column newsletter) will not extract as a single clean table.

- Scanned PDFs (image-only, no embedded text) won't extract anything — OCR them first.

- Always check the preview — heuristic column detection can merge or split columns on unusual layouts.

Limitations

- This is a position heuristic, not true table-structure detection — it can't reliably handle merged cells, nested tables, or rotated text.

- Tables that span multiple pages are extracted as separate sheets, not stitched into one continuous table.

- Very dense or tightly kerned text can throw off column-gap detection.

FAQ

How does the extraction work?

It reads the exact X/Y position of every word on the page, groups words into rows by closeness in vertical position, then detects column boundaries from the horizontal gaps that repeat across rows — no real table-grid data exists in a PDF, so this is inherently a best-effort reconstruction.

Why don't my columns line up correctly?

Uneven spacing, merged cells in the source, or a non-tabular layout can confuse the column-gap heuristic. Try a PDF with more consistent alignment, or expect to touch up the result in Excel.

Does this work on scanned PDFs?

No — scanned pages have no embedded text layer to read positions from. OCR the PDF first with a tool that adds a text layer.

Is anything uploaded?

No. The PDF is processed entirely in your browser; it's only sent anywhere when you choose to download the result.