PDF to Excel Converter — extract tables with OCR for scanned pages, right in your browser
Extract tables using position-based column detection, recognize numeric values including currency amounts and percentages, and export the result as .xlsx or .csv.
Position-based layout detection: Columns are identified from the horizontal positions of text elements on the page. Merged cells, rotated text, and charts fall outside what position detection can reconstruct.
Convert PDF to Excel
Upload a PDF, adjust detection options if needed, then download your spreadsheet.
Browser-Based Processing
Your PDF is processed in your browser tab; the document itself is not sent to an external conversion server.
Drop your PDF here
or browse files
Table preview
Export settings
Used as prefix: "Page 1", "Page 2", ...
Position-based table detection, with clear limits
Knowing how detection works helps you get better results and understand when manual adjustments are beneficial.
What it does
- Detects columns from the horizontal gaps between text on the page with configurable sensitivity.
- Runs OCR via Tesseract on scanned pages, using word-level position data for column detection.
- Converts numeric strings to numeric cells — handling thousands separators, currency symbols, parenthetical negatives, and percentages while keeping codes with leading zeros as text.
- Allows instant re-tuning after conversion — column mode, gap sensitivity, and sheet options recompute the preview immediately.
- Prompts for a password on protected PDFs in your browser.
Where it stops short
- Merged cells in the original table are not reconstructed as merged in the spreadsheet.
- Rotated text and charts are not detected or preserved.
- Irregular spacing can affect column detection — gap sensitivity and column mode are designed to adjust for this.
From PDF table to editable spreadsheet
Upload, convert, then adjust the result live before downloading.
Upload your PDF
Drop your PDF into the upload area, or click Select PDF File. If it is password protected, you'll be prompted for the password.
Set detection options, then convert
Choose column mode, gap sensitivity, and a page range if you don't need the whole document, then click Convert to Excel.
Review and adjust the preview
If columns look merged or split unexpectedly, adjust gap sensitivity or column mode — the preview updates instantly without re-reading the PDF.
Choose format and download
Select .xlsx or .csv, set sheet naming and styling options, then download.
Supported layouts and common use cases
Guidelines on what types of documents convert smoothly and where manual review helps.
| Use case | How it handles the layout | Suitability |
|---|---|---|
| A bank or financial statement | Works best when dates, descriptions, and amounts are arranged as consistent text columns. | Works well with clean layouts |
| A price list or invoice table | Works well when item names, quantities, and amounts have consistent spacing. | Works well with clean layouts |
| A scanned data table | OCR can provide word positions for scanned pages, after which position-based detection is applied. | Suitable with review |
| A table with merged header cells | Merged cells aren't reconstructed — the header row may need manual adjustment in Excel. | Needs review |
| A table with charts or embedded graphics | Only text is extracted — charts and images are not converted. | Not supported |
| Rotated or sideways table text | Column detection assumes standard horizontal text orientation. | Not supported |
How column detection works
The converter uses text positions and horizontal spacing to infer table columns.
Group text into lines by vertical position
Every piece of text on the page has position coordinates. Text sitting at approximately the same vertical height is grouped into a line in reading order.
Merge nearby text into cells based on gap size
Within a line, text separated by small gaps is joined into a single cell. A gap wider than the chosen sensitivity threshold starts a new cell.
Find column boundaries from aligned cells
Lines with two or more cells are evaluated to find where cell edges cluster together vertically across multiple lines, defining column boundaries.
Place each cell into its matching column
Each detected cell is assigned to the column it overlaps most. If it does not overlap a detected column, the nearest column is used instead, reducing the chance of losing extracted text.
When something doesn't look right
Common layout behaviors and quick adjustments.
Two columns got merged into one
The gap between columns was narrower than the current sensitivity threshold.
Switch Gap sensitivity to Tight and the preview updates immediately.
One column split into several
Word spacing within that column was wide enough to be treated as a column separator.
Switch Gap sensitivity to Loose to require wider spacing before splitting.
A number remained as text
It may have leading zeros (kept as text on purpose for IDs and postal codes) or exceed Excel's 15-digit precision limit.
This is intentional for code-like values. Ensure the "Detect numbers" checkbox is enabled.
"Incorrect password" message
PDF passwords are case-sensitive, and owner/permissions passwords cannot unlock reading access.
Check capitalization and confirm you are entering the user/open password.
Page range change did not update preview
Changing page range or OCR options requires re-reading the PDF source pages.
Click "Convert to Excel" again after updating the page range.
Header appears multiple times after merging
Header removal drops rows that match the first page's header text exactly. If OCR produces a slight character variation, it remains to avoid deleting data.
Delete the extra row manually in Excel, or re-run with text layer if available.
Columns merged
Why: Gap narrower than current sensitivity.
Fix: Switch to Tight.
Column split into several
Why: Wide word spacing.
Fix: Switch to Loose.
Number stayed as text
Why: Leading zeros or over 15 digits.
Fix: Intentional for IDs and codes.
"Incorrect password"
Why: Case-sensitive, or wrong password type.
Fix: Confirm exact open password.
Page range change did nothing
Why: Requires re-extracting PDF pages.
Fix: Click Convert again.
Header row appears more than once
Why: Slight character difference prevented exact match.
Fix: Delete the extra row manually in Excel.
Getting clean table extractions
Practical tips for common document layouts.
Start with Normal sensitivity, then adjust
Convert once, check the preview, and switch to Tight or Loose if columns need grouping adjustments.
Use a page range for long documents
If the table only spans a few pages, converting just those avoids unnecessary processing time.
Match OCR language to your document
Tesseract accuracy improves when the selected language matches the text in the scanned page.
Use "Merge all pages" for multi-page tables
Combined with "Remove repeated headers," this can reassemble a multi-page table while removing repeated headers that match the first page exactly.
Review numeric columns before formulas
Spot-check cells in the preview — right-aligned cells with green highlights indicate recognized numeric values.
Keep the original PDF for complex tables
Merged cells and graphics require manual layout recreation in Excel — the source PDF serves as an easy visual reference.
PDF to Excel FAQs
Answers to common questions about table detection, OCR, and formats.
How does column detection actually work?
It measures the horizontal gaps between text on each line and groups text into columns based on where those gaps line up across multiple lines. This is a heuristic based on layout geometry, so unusual layouts, merged cells, and multi-line cells can require adjustment. Gap sensitivity and column mode options are available to tune detection to your specific document.
Will scanned PDFs work?
Yes, using browser-based OCR (Tesseract) that runs locally in your browser, with word-level position data used for column detection just like text-layer pages.
Does it recognize numbers, currency, and percentages?
Yes. Values are recognized and converted into numeric cells, handling thousands separators, currency symbols, parenthetical negatives, and percentages — while keeping ID numbers with leading zeros as text to preserve their format.
What happens with a password-protected PDF?
You will be prompted to enter the document password. Once unlocked in your browser, table extraction proceeds normally.
Are merged cells, rotated text, or charts preserved?
No. These fall outside what position-based column detection can reconstruct, so charts, rotated text, and cell merging are not preserved.
My table spans several pages — will the header repeat throughout?
Not if you choose "Merge all pages" with "Remove repeated headers" enabled — it keeps the first header row and appends data rows from subsequent pages when their headers match the first page.
Are my PDF files uploaded to a conversion server?
No. The conversion, OCR, and spreadsheet generation happen entirely in your browser, so the PDF is not sent to an external conversion server.