Choose a PDF and its tables are reconstructed from the position of the text on each page — rows and columns detected from alignment — then shown on screen and downloadable as CSV for Excel or Google Sheets. By default only real tables are extracted (pages of ordinary prose are skipped, not dumped into your CSV); switch Output to "Everything" if you want every page as a grid. Best effort: clean, aligned tables extract well; merged cells, borderless layouts and scanned pages (no text layer) may need tidying. Runs fully in your browser — nothing is uploaded.
Liberate the numbers trapped in a PDF table: out as a tidy CSV your spreadsheet actually understands.
100% free · No sign-up · No daily limits · No watermarks · Processed in your browser — files never leave your device
Extract tables from a PDF into a real CSV file (and an on-screen table you can check first): text is grouped by position into rows and columns — best-effort for clean, text-based tables.
A 12-page price list PDF becomes a spreadsheet in a minute: item, pack size and price line up in columns, ready to sort and total — work that would have been an afternoon of retyping.
The tool reads the PDF’s positioned text in your browser and reconstructs the grid geometrically: words sharing a baseline become a row; vertical gaps between text clusters become column boundaries. That is how most “PDF table extractors” work without AI — deterministic geometry, not document understanding — and the CSV is standard comma-separated text any spreadsheet opens.
Their spacing was too tight for the geometry to see a boundary. Nothing was lost — the CSV keeps the text — but you will need to split that column in your spreadsheet.
No — scans have no text layer to group. Run OCR PDF first to add one, then extract (accepting OCR’s errors ride along).
All of them: every page is analysed, each detected table is previewed on screen, and the tables are written in page order — combined into one CSV, or as one CSV per table in a ZIP if you pick that Output. Pages with no table structure are skipped, and the done message counts them.
CSV opens natively in Excel and every spreadsheet, and “Save as” converts it in one step — CSV is the universal handover format, so that is what we produce.
No. Extraction runs entirely in your browser.