PDFCellsConvert an invoice PDF to Excel
Open the invoice in PDFCells and download the line-item table as Excel. On the way, it checks that quantity × unit price equals the line total and that subtotal plus tax equals the invoice total, and flags lines that don’t.
Steps
- Go to pdfcells.com and choose the PDF, or drop it on the page. Drop several files to open them as a batch. A password-protected PDF asks for its password. Photos and scans of invoices open too (JPG, PNG, WebP, GIF, HEIC, TIFF).
- Every page is pre-checked and gets a read method: PDF text for pages with a text layer, OCR for scans, or AI for pages you choose.
- Each table opens in an editable grid next to the page. Totals that don't add up, balances that don't tie, text in a number column and cells OCR was unsure of are flagged. Fix any cell, merge or split columns, or undo. If the invoice has text around the table you don’t want, mark the area to keep or ignore.
- Download Excel. With summary sheets on, a By item sheet totals each item with live formulas.
Checks on an invoice
| Check | What must hold |
|---|---|
| Line totals | Quantity × unit price = line total, once most rows of the table fit that pattern. Rows with a quantity of 1 can’t establish the pattern on their own, but are checked once it is established. |
| Subtotal and tax tie to total | Subtotal + recognised tax lines = total. Shipping and other charges aren’t counted, so an invoice with them can show a false warning on its total. |
| Footed | A row labelled Total equals the rows above it. |
| Number columns | Text in a number column, or a cell OCR read with under 70% confidence, is flagged. |
Numbers from any locale
European invoices that write 1.234,56 and US ones that write 1,234.56 both become the number 1234.56. Currency symbols and codes are stripped from the value and the column keeps a currency format.
Many invoices at once
Drop up to 200 PDFs to open a batch. Structural fixes you make while reviewing the first file of a layout (areas, merged or split columns) are replayed on the other files of that layout, and the batch exports one workbook with a Source file column so every line traces back to its invoice. One file can have up to 300 pages. A batch holds up to 200 files and 2,000 pages in all.
PDFCells extracts tables. It doesn’t pull header fields such as vendor name, invoice number or due date into named fields; if you need those as structured data across thousands of invoices, a template-based tool such as Docparser or an AI extraction platform fits better.
Questions
Can it read scanned or photographed invoices?
Yes. Scans and photos are read with free OCR in your browser, or with AI for skewed or low-quality pages. A misread price usually breaks the line-total check, which flags it.
Does it extract the invoice number and vendor?
Not as separate fields. PDFCells converts tables; the header text is available on the page but isn’t mapped to named fields.
Is it free?
Reading the PDF text and OCR are free, need no account and run in your browser. AI reading is optional: signed-in users get 25 free AI pages a day on the Fast model, then prepaid credits at Anthropic's token cost plus 15% (packs of $10, $25, $50), or no markup with your own Anthropic API key.