PDF to Excel
Tables from a PDF, in a sheet you can put a formula in.
This tool uploads your file for conversion and deletes the temporary copy when the request finishes. What happens to your files
This tool uploads your file to a server to convert it.
Your file is uploaded to a server for this conversion. It is written to a temporary directory, used, and deleted before the response is sent — within seconds, in the same request. It is never written to a database, an object store, or a queue, and it is never logged.
What survives: Each table becomes its own sheet with a frozen header, and figures arrive as numbers you can sum. A PDF has no concept of a cell, so borderless tables are inferred and can be wrong. Check the figures against the original.
A quick browser check runs when you start.
How it works
- Read the live upload and retention statement.
- Choose a PDF containing tables and start conversion.
- Download the XLSX and verify every important figure.
Key benefit
Places detected tables on sheets and stores recognized numeric values as numbers.
Supported files and key limits
Input: PDF. Output: XLSX.
Borderless tables are inferred, tables the converter is unsure of are left out, and amounts with currency symbols stay as text. Check every figure.
The failure nobody notices: numbers stored as text
A converted sheet where 125000 is stored as the text “125000” looks completely normal, and the column may even be right-aligned. Then you select it and the total reads zero, or a formula elsewhere returns a plausible wrong number that nobody checks. PDFOak writes plain numbers as numbers, including values with thousands separators and negative signs, so those columns sum as soon as they land.
Some values are handled differently. Percentages become decimals, so 12% is stored as 0.12. Amounts with a currency symbol or brackets, such as $1,250 or (1,250), stay as text and will not sum until you clean them. Dates stay as written, which avoids a silent day/month swap. Leading zeros on numeric-looking codes are dropped, so check postcodes and account numbers.
A PDF has no idea what a cell is
Nothing in a PDF records that four numbers form a row. Ruled tables give their structure away, and extraction is reliable. Where there are no ruling lines, which is common in annual reports and schedules, columns are inferred from spacing, and the inference is sometimes wrong.
If the converter is not confident about a table, it leaves that table out instead of guessing which values belong in each cell. A document that mixes ruled and borderless tables can also come back with only the ruled ones, and a document with no detectable tables produces an empty workbook. Check that every table arrived as well as every figure.
How the workbook is laid out
Each table becomes its own sheet, named Table 1, Table 2 and so on, with the header row frozen and column widths set from the content. Tables are not stacked on one sheet, because stacking breaks every formula reference the moment a table above changes height.
Scanned or protected PDFs
A scanned table has no text to extract, only an image of one. Run OCR PDF first to add a text layer, then convert. OCR is good on numerals but not perfect, so a misread digit in a financial table is the error to look for. The converter also refuses password-protected PDFs: remove the password with Unlock PDF first.
This one uploads your file
Finding table structure takes more than a browser tab can do, so PDFOak uploads the file, converts it, and deletes the temporary copy when the request finishes. The privacy page lists which PDFOak tools upload.