Python PDF Table Extraction: Complete Tutorial

Last updated: March 2026 • 10 min read

Automate PDF table extraction with Python. This tutorial covers three popular libraries with working code examples.

Setup

pip install camelot-py[cv] tabula-py pdfplumber pandas openpyxl

Note: Camelot requires Ghostscript. Tabula requires Java.

Method 1: Camelot

Best for tables with clear borders (lattice) or consistent spacing (stream).

import camelot

# For bordered tables
tables = camelot.read_pdf('document.pdf', pages='1', flavor='lattice')

# For borderless tables
tables = camelot.read_pdf('document.pdf', pages='1', flavor='stream')

# Export to Excel
tables[0].to_excel('output.xlsx')

# Get as pandas DataFrame
df = tables[0].df

Method 2: Tabula-py

Java-based engine, good general-purpose extraction.

import tabula

# Extract all tables from all pages
tables = tabula.read_pdf('document.pdf', pages='all')

# Returns list of DataFrames
for i, table in enumerate(tables):
    table.to_excel(f'table_{i}.xlsx')

# Specify area (top, left, height, width in points)
tables = tabula.read_pdf('document.pdf', 
                         area=[0, 0, 500, 800],
                         pages='1')

Method 3: pdfplumber

Pure Python, no external dependencies. Good for simple tables.

import pdfplumber
import pandas as pd

with pdfplumber.open('document.pdf') as pdf:
    page = pdf.pages[0]
    table = page.extract_table()
    
    # Convert to DataFrame
    df = pd.DataFrame(table[1:], columns=table[0])
    df.to_excel('output.xlsx', index=False)

Which Library to Choose?

• Camelot: Best accuracy, especially for bordered tables
• Tabula-py: Good balance of features and ease of use
• pdfplumber: Simplest setup, no Java/Ghostscript needed

Handling Edge Cases

# Multiple pages
tables = camelot.read_pdf('doc.pdf', pages='1-5,8')

# Scanned PDFs (need OCR first)
# Use pytesseract or cloud OCR, then extract from text layer

# Tables spanning pages
# Extract separately, then concatenate DataFrames
df_combined = pd.concat([t.df for t in tables])