How to Import Questions and Answers from PDFs and Photos
Reliable import depends as much on choosing the right path as on the extraction model. A separate answer booklet and one geometry diagram should not enter the same way.
Identify the shape of the source first and the first note takes far less correction.
Match the entry point to the source
- Questions and answers in one PDF — upload the whole file together.
- Question PDF and answer PDF are separate — identify each source separately.
- Several photos — put them in question-number order first.
- One diagram-heavy problem — crop it as an image so the figure remains.
- One problem from class — direct entry is usually fastest.
Question numbers are the join
When questions and answers live apart, their numbers make the pair. Fix shuffled pages or cropped numbers before upload.
Check whether parts (a) and (b) should stay under one question or become separate entries.
Keep the picture when layout carries meaning
Text extraction may turn a picture into a short description. Coordinates, geometry, circuits, and maps are better kept through image cropping.
A practical note can use searchable text for the prompt and retain an image for the figure.
Review four things before saving
- Question boundaries
- Question-to-answer numbering
- Answer versus explanation fields
- Missing information in figures or formulas
Do not polish every sentence. The standard is whether you can solve the problem again correctly.
Change paths when the source resists extraction
Overlapping handwriting and irregular layouts can be faster to crop or enter directly.
AI extraction may use credits. Change the input method instead of repeatedly running the same unsuitable source.