Medical Question Answering on MedQuAD-style Complete Benchmark
91.1MedQuAD ScoreMed-Gemini
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Med-Gemini2026.01 | 91.1 | — | |
| Med-PaLM 22026.01 | 86.5 | — | |
| Knowledge-Embed Model2026.01 | 82.92 | — | |
| GPT-42026.01 | 71.07 | — | |
| Med-PaLM (original)2026.01 | 67.6 | — | |
| MeddollinaRegime=Baseline 4 - CCI reference instantiation, Protocol=governance-first clinical intelligence2026.01 | — | 100 |