Which medical AI would you trust?

You'll compare two anonymized models shown as benchmark-strength bars and pick the one you'd rely on. This calibrates how we weight medical AI benchmarks.

It measures how people weight benchmark dimensions. It is not a clinical-outcomes study.