AI-PDF-Renamer

🔍 Automatically extract metadata from PDF files and rename them using AI

This tool uses Google's Gemini AI to analyze PDF documents, extract key metadata (title, author, publication year), and automatically rename files in a standardized format for better organization.

🚀 Features

• AI-Powered Extraction: Uses Google Gemini AI to read PDF content and extract metadata
• Multiple AI Models: Support for Flash (gemini-2.5-flash-preview-05-20) and Pro (gemini-2.5-pro-preview-06-05) models
• Smart Renaming: Automatically renames files in format: YEAR - AUTHOR - TITLE.pdf
• Batch Processing: Process entire directories of PDF files at once
• Non-Destructive: Creates renamed copies while preserving original files
• Rate Limiting: Respects API limits with intelligent retry logic

📦 Installation

The run.sh script handles environment setup automatically, including virtual environment creation and dependency installation.

📖 Usage

# Setup: Get API key from https://aistudio.google.com/app/apikey
echo "GEMINI_API_KEY=your-actual-api-key-here" > .env

# Basic usage - process PDFs and rename them
./run.sh ./documents ./organized

# Extract metadata only (no file copying, saves JSON to source dir)
./run.sh ./documents ./results --no-copy

# Process only first page (faster/cheaper)
./run.sh ./documents ./organized --max-pages 1

# Use Pro model (gemini-2.5-pro-preview-06-05) for better accuracy
./run.sh ./documents ./organized --model pro

# Automation mode (skip confirmations)
./run.sh ./documents ./organized --force

# Combine options
./run.sh ./papers ./organized --max-pages 1 --force

📊 Example Results

Input: sample.pdf
Output: 2015 - André Koch Torres Assis - A new method for inductance calculation.pdf

Results are also saved to pdf_metadata_results.json with detailed metadata for each processed file.