Detect tables in a PDF and export them as CSV, Excel (XLSX) or JSON - bordered or borderless. Locally in your browser, no upload.
Running locally on your device ...
Running locally on your device ...
0%
Your files never left your device
No. Everything runs in your browser - your file never leaves your device. How this is verifiable
Numbers often sit exactly where you cannot work with them: in PDF tables. Invoices, reports, bank statements, studies - anyone wanting the values in Excel, a database or a script has so far retyped them or fought columns that fall apart on copy-paste. This tool detects tables in the PDF automatically and exports them with structure intact: as CSV for spreadsheets and imports, as an Excel workbook (XLSX) with one sheet per table, or as JSON with coordinates for further processing in code.
Detection uses the same two techniques the well-known specialist tools of this class rely on: tables with ruling lines are recognised from the lines actually drawn on the page (grid mode) - the lines form the cell grid, and each piece of text is assigned to its cell. Borderless tables are recognised from alignment (spacing mode): text blocks that sit at the same column positions across several lines form the columns, and the gaps that stay clear become the column boundaries. The automatic setting checks for lines first and falls back to spacing analysis; both modes can also be forced. Multiple tables and multi-page PDFs are supported - every detected table is exported separately.
Honest and important: detection reads the text layer of the PDF. A scanned PDF, or one made entirely of images, has no such layer - there the tool says so openly instead of producing empty or invented tables. Convert scans to searchable text with "PDF OCR" first, then extract the tables. Complex layouts have limits too: merged cells, nested tables or cells with many line breaks are approximated, not reconstructed pixel-perfectly. Everything runs locally in the browser via pdf.js; your file never leaves your device.
| Input formats | |
|---|---|
| Auto-conversion | PNG, JPG, GIF, BMP, TIFF, ZIP, DOCX |
| Output formats | CSV, XLSX, JSON |
| Batch processing | No |
| Processing | Locally in your browser (JavaScript) |
| File upload | None |
Limitations: Detects tables via ruling lines (grid) or text alignment (spacing) and exports them as CSV (one file per table, several as ZIP), XLSX (one sheet per table) or JSON (with page, cell and position data in PDF points). Requires a real text layer - convert scanned or image-only PDFs with "PDF OCR" first, otherwise an honest notice appears instead of an empty table. Merged cells, nested tables and heavily designed layouts are approximated; the Excel export only converts unambiguous numbers (dot decimal) into number cells. Open password-protected PDFs with "Unlock PDF" first.
With two techniques: for bordered tables the drawn lines form the cell grid (grid mode). For borderless tables, columns are recognised from text alignment and the gaps that stay clear (spacing mode). The automatic setting picks the right technique itself.
A scan is only image pixels and has no text layer to read. Convert the PDF to searchable text with "PDF OCR" first, then extract the tables.
Yes. All pages are scanned and every detected table is exported separately: as its own CSV file (several bundled as ZIP), as its own sheet in one Excel file, or as an entry in the JSON.
The full structure: per table its page, detection mode and position, and per cell its text and bounding box in PDF points (origin bottom left) - ideal for scripts and data pipelines.
A cell merged across several columns has no unique position in the grid; its content lands in one of the involved cells and the others stay empty. That is an honest approximation - even specialised desktop tools do not solve this flawlessly.
No. Detection and export run entirely locally in the browser via pdf.js; the file never leaves your device.
Yes, on any device with a modern browser - Windows, Mac, Linux, Chromebook or a smartphone. Very large PDFs just take more time and memory on mobile devices.
PDF to text · PDF OCR (searchable) · XLSX to CSV · CSV to XLSX · CSV to JSON · PDF to Markdown