Medical Instruction Following on MedMTbench
0.6375Pass@1MedXIAOHE
Evaluation Results
| Method | Links | |
|---|---|---|
| MedXIAOHEMode=Thinking mode, Decoding=Greedy2026.02 | 0.6375 | |
| GPT-5.2 ThinkingMode=Thinking mode, Decoding=Greedy2026.02 | 0.513 | |
| Gemini 3.0 ProDecoding=Greedy2026.02 | 0.498 | |
| Gemini 2.5 ProDecoding=Greedy2026.02 | 0.45 |