I see a lot of people defaulting to searchable PDFs for OCR, but depending on your use case (like feeding LLMs or long-term archiving), JSON or PDF/A might be m

I see a lot of people defaulting to searchable PDFs for OCR, but depending on your use case (like feeding LLMs or long-term archiving), JSON or PDF/A might be much better. I put together a comparison of TXT, PDF, PDF/A, XML, and JSON to show the trade-offs in size, structure, and compatibility. Hope this helps anyone working on digitization!