Model evaluation
ReMedQA closed
| Model | Strict ClosedAvg |
|---|---|
| Moon 1 | 89.56% |
| GPT-6 Astra | 87.39% |
| Claude Opus 5 | 89.38% |
Strict ClosedAvg measures weighted accuracy across six ReMedQA medical multiple-choice formats.
| Model | Strict ClosedAvg |
|---|---|
| Moon 1 | 89.56% |
| GPT-6 Astra | 87.39% |
| Claude Opus 5 | 89.38% |
Strict ClosedAvg measures weighted accuracy across six ReMedQA medical multiple-choice formats.