Shortexam Credit is the unit used to measure AI resource consumption on the platform.

A token is a chunk of text processed by an AI model. Input and output are both measured in tokens.

In question paper generation, selected chapter content counts as input tokens. Generated questions and model reasoning count toward output and thinking tokens.

PDF processing is page-based. Processing 2 pages consumes 1 credit.

Basic uses a medium-sized model with limited thinking budget. Standard uses a larger model with higher thinking budget. Premium uses the most capable frontier model with the largest thinking budget.