Extract Tables
Detect tables in a PDF and export them as CSV, JSON or Markdown with rows and columns intact, ready for a script or a spreadsheet.
Charger un fichier PDF
À propos de cet outil
Table data locked in a PDF is the most common data-entry trap there is. Copying and pasting a table into a spreadsheet produces a single column of mush, because PDFs store where each piece of text sits rather than which cell it belongs to.
This detects table structure - from ruling lines where they exist, and from column alignment where they do not - and exports real rows and columns. Three output formats cover the usual destinations: CSV for spreadsheets, JSON for scripts, and Markdown for documentation.
Every detected table is previewed before export, so you can confirm the column boundaries landed correctly rather than discovering a misalignment after you have built something on top of the data.
Comment utiliser
Load the PDF
Drop in the file. Pages are analysed for table structures.
Check the detected tables
Each is previewed with its rows and columns so you can verify the boundaries.
Choose the export format
CSV for spreadsheets, JSON for code, Markdown for docs.
Download
Save the extracted data.
Cas d'utilisation
Financial data into a model
Statement tables exported as CSV and dropped straight into a spreadsheet.
Automated data collection
JSON output feeding a script that processes many reports on a schedule.
Documentation
Markdown tables pasted directly into a wiki page or README.
Foire aux questions
What makes a table detectable?
Ruling lines, or consistent column alignment. Tables held together only by irregular whitespace are the hard case.
How are merged cells handled?
A merged cell is assigned to its first position and the remaining cells are left empty. Check the preview when your table has spanning headers.
Does it work on scanned PDFs?
Not directly. OCR the document first so there is text to analyse.
Should I use this or PDF to Excel?
PDF to Excel produces a formatted spreadsheet. This gives you raw CSV, JSON or Markdown, which is better for scripts and version control.