PDF support

• Claude can read and process content directly from PDF files, not just images
• To read a PDF, use nearly identical code to reading images, with key modifications:
• Change file reference from image to PDF filename
• Change variable name from image_bytes to file_bytes (or similar) for clarity
• Change type from image to document
• Change media type from image/PNG to application/PDF

• PDF files must be placed in the same directory as your notebook/script

• Claude can extract multiple types of content from PDFs:
• Text content
• Images and charts
• Tables
• Other visual elements

• Claude functions as a versatile tool for extracting various kinds of information from PDF documents in a single workflow

Takeaways

• Claude can read and process PDF files directly using nearly identical code to image reading, with changes to file reference, variable name, type (document), and media type (application/PDF)
• PDF files must be in the same directory as your notebook/script
• Claude can extract multiple content types from PDFs including text, images, charts, and tables in a single workflow