How to convert a PDF to Excel without uploading it
Quick answer
Open ShushPDF's PDF to Excel and tap Choose PDF. The tables are found on your own device. Pick one in the Table list, check the preview, then tap Save XLSX or Save CSV. Nothing is uploaded. It works on PDFs that have real text, such as downloaded statements and reports, not on scans.
Bank statements, payslips and invoices are often the PDFs people want in a spreadsheet, and they are also the ones you should not upload to a stranger's server. This guide shows how to get a PDF table into Excel while the file stays on your phone or computer, how to check that for yourself, and what to look out for.
How to convert a PDF to Excel with ShushPDF
- Open PDF to Excel in any modern browser.
- Tap Choose PDF and pick your file. The page reads it and says, for example, "Found 154 tables on 71 pages."
- Pick a table in the Table list. Each one shows its page and size, like "Table 39: page 12, 7 × 10" (rows × columns).
- Check the preview. It shows the first 30 rows.
- Tap Save XLSX for Excel, Google Sheets or Numbers, or Save CSV for other programs.
Each file holds one table. For several tables, pick the next one and save again. If the PDF has a password, type it when asked; it is used only on your device.
How to check that nothing is uploaded
- Airplane mode: convert one PDF, turn on airplane mode, reload the page and convert again. It still works, because there is no server doing the work.
- Network tab (on a computer): press F12, open Network, then convert a PDF. In our test with a 71-page PDF, the only two requests while converting fetched the page's own PDF reader. None carried the file.
Our guide on checking whether an online PDF tool uploads your file explains both tests in more detail.
What we found with a real tax table
On 13 November 2026 we converted IRS Publication 15-T, a 71-page PDF full of rate tables. We used the PDF only as a test file, so nothing here is tax advice.
| What we checked | Result |
|---|---|
| Tables found | 154 in 1.4 s |
| Table 39 (page 12) | 7 rows × 10 columns, all 70 cells right |
| Numbers in the XLSX | "$19,300" → 19300, "10%" → 0.1 |
| First row of each rate table | Moved into the table above |
Check the first rows
Above each rate table, the IRS prints its column headings over several lines. PDF to Excel saw that block as its own table, and the first row of numbers ("$0, $19,300, $0.00, 0%, $0") went into it, squeezed into one cell. The rest of the table was right. So before you save, compare the first and last rows of the preview with the PDF. If a row is missing, look in the table just before it in the Table list, and copy that row across in Excel.
Headings stacked over two or three lines can also come out over several rows. That is easy to fix in Excel: delete the extra rows and type one heading per column.
Scanned PDFs
A scan is a picture of a page, with no text inside. PDF to Excel then says "No text found in this PDF" and that it is probably a scan. OCR PDF can add the words of a scan as hidden text, but we have not tested whether the columns of a scanned table then come out right, so check the preview carefully.
FAQ
Is my PDF uploaded?
No. The tables are found and the Excel or CSV file is made inside your browser. You can check with airplane mode or your browser's Network tab, as shown above.
Should I save XLSX or CSV?
XLSX for Excel, Google Sheets or Numbers: amounts become numbers you can add up. CSV for importing into another program: every cell stays exactly as printed in the PDF.
Why is a row missing from my table?
It probably went into the table before it, as the first row of each IRS rate table did in our test. Pick the previous table in the Table list and look at its last rows.
Can I convert a bank statement?
Yes, if it is a PDF you downloaded from your bank, as those have real text. The PDF to Excel page lists our tests with two sample statements. A photo or scan of a paper statement is a picture, so it does not work.
How this guide was made: drafted with AI help, tested with scripted browser runs with ShushPDF on 13 November 2026, and reviewed by the ShushPDF team before publishing.