A question bank that reads your reasoning.
Static banks give everyone the same next question. Medutors chooses yours from how you reasoned on the last ones: the band you are in, the clusters you break on and the concepts due for review.
Four signals choose your next question
Adaptation is based on reasoning outcomes, not only correct or incorrect. A right answer with a missed key clue still counts as a partial break.
Difficulty band
Each question carries a difficulty band. Consistent accuracy in a band moves you up; repeated reasoning breaks move you back to consolidate.
Weak-cluster weighting
Clusters from your diagnostics report get a larger share of the next set until the breakdown rate falls.
Spaced return
Concepts you have recently fixed come back at widening intervals, so improvement is tested rather than assumed.
Cross-subject mixing
Sets mix subjects the way real papers do, so you learn to recognise a pharmacology question hiding in a surgical stem.
A typical adaptive set, broken down
The proportions shift with your profile. Early on, coverage dominates. Once weak clusters are clear, they take the largest share until they stop being weak.
All 19 subjects, mapped into concepts
Questions are tagged by subject and by the concept they test. That second tag is what lets a renal acid-base error and an endocrine acid-base error count as the same weakness.