Reconstruct rows and columns from selectable PDF text, review every cell, and export CSV or XLSX without uploading the document. Scanned tables should be processed with local OCR first.
100% private: everything runs on your computer, your files never leave your device.
PDF to Excel finds the tables in a PDF and turns them into rows and columns you can edit, on your own computer. It reads the position of every piece of text on the page and rebuilds each table from the way the words line up, so a table does not need borders to be found. Before anything is exported, you see each table in a preview where every cell can be corrected.
You can export an Excel workbook or a CSV file. The workbook has a sheet called Original, with every page of the PDF as a picture so you can compare the data with the source, and a data sheet with all the detected tables, where numbers, percentages, amounts and dates are real values. CSV exports the selected table as plain data for any program.
The look of the original tables, with their colours, fonts and borders, is kept only as a picture in the Original sheet. The data sheet has its own simple format, made to be sorted, filtered and reused.
The text, its positions and the page pictures are read, and the workbook is written, by code running in your browser tab. OCR, when you ask for it, also runs locally with files served from OléPDF's own domain. Nothing is uploaded and no page is sent to a conversion service.
That is what makes it usable with bank statements, invoices, payroll reports or price lists, the documents where tables usually live.
Yes, the free tools include 5 PDFs a day in total. Extracting the tables of one PDF at a time is free, with no account. Processing several PDFs in a single run requires the 7-Day Pass or Pro.
Either the PDF has no selectable text, which is the case with scans, or its content is not aligned in rows and columns. For scans, press Retry with local OCR. The XLSX can still be exported with every page in the Original sheet.
Check the preview before exporting. The detection works from text positions, so merged cells, wrapped text or unusual layouts can put a value in the wrong row or column. Any cell can be edited there.
Yes, in the XLSX. Numbers, percentages, amounts in euros, dollars or pounds, and dates written day first become real values. A CSV file has no types, so there every value is plain text.
Yes, with local OCR in Spanish and English, which runs only on pages without text. Review the recognised cells carefully, especially with low quality scans.
No. Detection, OCR and export run on your computer inside the browser tab, and the PDF never reaches a server.