Why did my table extract with jumbled columns?

The most likely cause is that the table in the PDF was built by positioning text manually rather than using a real table structure. Visually it looks like a table, but the PDF stores it as individual text elements at specific coordinates. Extraction has nothing to follow and the result reflects the underlying disorder. This is not fixable by re-running the extraction.

Can I extract tables from a scanned PDF?

Yes. Scanned PDFs are detected automatically and processed with OCR before extraction, so the recognized text is then analyzed for table structure the same way it is for a normal PDF. Accuracy depends on the table itself: scanned tables with real, visible grid lines and clear cell alignment extract well, while tables that rely on faint or inconsistent spacing rather than actual borders are harder to reconstruct correctly once the text has gone through OCR, the same limitation that already applies to non-scanned PDFs with spacing-only tables.

What if a table spans multiple pages?

Multi-page tables are detected where the column structure remains consistent across pages. If the column headers repeat on each page, they may also repeat in the output. Clean that up in the spreadsheet after extraction.

How is this different from PDF to Excel?

PDF to Excel converts the whole document (text, layout, and tables) into a single spreadsheet sheet. Extract Tables focuses specifically on table data: each detected table is placed on its own labelled sheet, making it easy to work with individual datasets without sifting through unrelated content.

What format is the output file?

The output is an .xlsx file. Each detected table is placed on a separate, labelled sheet. The file is fully compatible with Microsoft Excel, Google Sheets, LibreOffice Calc, and Apple Numbers.

What is the file size limit?

Each PDF can be up to 50MB on the free tier.

Extract Tables from PDF Online, Even Scanned PDFs

Pull every table from your PDF into a structured Excel spreadsheet, including scanned PDFs recognized automatically with free, built-in OCR. Each detected table lands on its own sheet, labelled and ready to use. No software, no signup required.

Drop your PDF here

or click to browse

Select PDF

Up to 50 MB

Enterprise-grade security

Files auto-delete

Free to get started

How to Extract Tables from PDF

1
Upload your PDF
Click the upload area or drag and drop your PDF. All PDF versions are supported up to 50MB.
2
Extract the tables
Click Extract Tables. PDFHaul detects every table in the document and maps rows, columns, and cell values into spreadsheet format.
3
Download your spreadsheet
Download the .xlsx file. Each table is on its own labelled sheet, ready to work with in Excel or Google Sheets.

Why use PDFHaul
to extract PDF tables?

One sheet per table

Each detected table is placed on its own labelled sheet in the output .xlsx file. No mixing of unrelated data. Every table is isolated and ready to use.

Table structure preserved

Rows, columns, and cell values are mapped accurately into spreadsheet cells. Spend less time reformatting and more time working with the data.

Free OCR for scanned PDFs

Scanned bank statements, invoices, and reports are recognized automatically and their tables extracted as real data, not left as an unusable picture. No separate OCR fee, no paid tier, no daily limit.

No watermarks, ever

Your extracted spreadsheet is clean. No branding, no watermarks, no trial stamps on your data.

Who extracts
tables from PDFs?

Structured data locked inside PDFs becomes immediately usable once extracted into a spreadsheet.

Financial analysts and accountants

Financial statements, audit reports, and balance sheets distributed as PDFs, including old paper statements that were scanned, often contain tables that need to be analysed in a spreadsheet. OCR handles the scanned ones automatically, so you extract once and work with the numbers directly.

Researchers and data teams

Government statistics, academic papers, and survey results frequently include data tables. Extract them into Excel for analysis, visualisation, and integration with other datasets.

Procurement and operations

Supplier price lists, inventory reports, and logistics data sent as PDFs can be extracted to spreadsheets for comparison, filtering, and import into business systems.

Legal and compliance teams

Contracts, regulatory filings, and audit reports often contain structured data tables. Extract them to Excel for record-keeping, tracking, and reporting workflows.

How table extraction works

Table extraction identifies structured tabular data within a PDF and outputs it as a spreadsheet. The accuracy depends on how the table was encoded in the original PDF.

PDFs created from Excel, Word, or database exports embed tables with clear column and row boundaries that extraction follows reliably. PDFs created by scanning a printed table, or by manually positioning text to look like a table, do not have underlying structure. The first type extracts cleanly. The second produces inconsistent results regardless of tool.

Multi-page PDFs are processed in full. Tables that span across pages are identified and combined where the column structure is consistent.

Scanned PDFs don't have a text layer at all: each page is a photograph of the table, not machine-readable characters. PDFHaul detects this automatically and runs the page through OCR (optical character recognition) first, recognizing the actual text and numbers, before handing the result to the same table-detection process used for any other PDF. This is included free, with no separate OCR fee or usage limit. Because OCR works from pixels rather than exact PDF coordinates, this works best on tables with real, visible grid lines; tables that rely purely on tight spacing to imply structure, with no borders at all, are harder to reconstruct correctly once the text has gone through recognition, the same limitation that already applies to non-scanned PDFs with spacing-only tables.

Frequently asked questions

Everything you need to know about extracting tables from PDFs

Yes. When a page has no real text layer, PDFHaul recognizes the text and numbers automatically using OCR (optical character recognition) before looking for table structure, so scanned invoices, statements, and reports can be extracted the same way as a native, text-based PDF. This is included free, with no separate OCR fee, paid tier, or daily limit.

From our blog

How to Extract Tables from PDF (2026)

Most people dealing with tables in a PDF do one of two things: take a screenshot or retype everything manually. Both work but neither is efficient when the table has dozens of rows or you need the data in a spreadsheet for analysis. What most people do not know is that free tools exist specifically for this problem.

6 min read