Text-Layer PDF vs Scanned PDF: What SmartSplitAI Accepts

A common question: "I have a contract as a PDF. Will it work?" The answer depends on the PDF type.

Text-Layer PDF — Works

If the PDF was created from a document (saved from Word or exported from Google Docs), it contains a text layer. SmartSplitAI processes these files.

How to check: open the PDF and try to select text with your mouse. If text highlights and copies — the file works.

Scanned PDF — Does Not Work

If the PDF came from a scanner or was photographed — it's an image, not text. There is no text layer. SmartSplitAI does not accept these files.

How to check: if the cursor draws a selection rectangle when you try to select text, and text doesn't highlight — it's a scan.

Why No OCR

OCR is a separate complex problem:

• Recognition errors are inevitable.
• One misread digit in a contract amount can fundamentally change the meaning.
• For contract analysis, accuracy is critical.

SmartSplitAI may add scan support in the future when OCR quality reaches an acceptable reliability threshold.

What to Do with a Scan

If you only have a scan, try to find the source document (DOCX, Google Doc). Or use PDF-to-DOCX conversion tools, but note: conversion quality may not be perfect.