Image-based quiz questions are frustrating because normal text extraction often fails. A chart, screenshot, PDF fragment or locked question can break a standard parser.

FocusScan lets you draw a box around the relevant area and sends that cropped context through OCR before AI analysis.

Try QuizSolver in practice

Open the safe demo, launch the extension panel and walk through sample questions without spending credits.

Open demo

Best use cases

Use FocusScan whenever the visible page structure does not contain enough text or when the important information is inside an image.

  • screenshots
  • charts and diagrams
  • math formulas
  • PDF fragments
  • blocked copy text

How to capture a useful region

Include the question, all answer options and any diagram needed to solve it. Avoid selecting the whole page because extra text can confuse the model.

Save the result for later

If the explanation is useful, save the question and add a short note. Image questions often repeat as concepts, even when the exact picture changes.