Nested-calculator calculation on MedCalc-Pro 1.0 (test)
87.22R-F1MedCalc-Pro Agent Framework
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MedCalc-Pro Agent FrameworkModel=gpt-oss-120b2026.07 | 87.22 | 93.5 | 88.48 | 75.8 | 87.1 | |
| ReActModel=gpt-oss-120b2026.07 | 80.62 | 83.87 | 66.42 | 11.74 | 77.41 | |
| MedCalc-Pro Agent FrameworkModel=GPT-5-mini2026.07 | 80 | 93.5 | 88.91 | 83.9 | 85.5 | |
| MedCalc-Pro Agent FrameworkModel=Qwen3-235B-A22B2026.07 | 78.26 | 87.1 | 81.05 | 75.81 | 80.65 | |
| ReActModel=GPT-5-mini2026.07 | 77.78 | 90.3 | 72.01 | 25.8 | 82.3 | |
| ReActModel=Qwen3-235B-A22B2026.07 | 74.02 | 75.81 | 58.45 | 19.35 | 56.45 | |
| ReActModel=MedResearcher-R1-32B2026.07 | 67.2 | 67.74 | 49.95 | 24.19 | 27.42 | |
| MedCalc-Pro Agent FrameworkModel=MedResearcher-R1-32B2026.07 | 61.54 | 64.52 | 57.27 | 56.45 | 91.94 | |
| MedCalc-Pro Agent FrameworkModel=DeepSeek-V3.12026.07 | 58.73 | 59.68 | 52.53 | 51.61 | 56.45 | |
| ReActModel=Llama-3.3-70B-Instruct2026.07 | 55.76 | 71 | 25.19 | 11.29 | 12.9 | |
| MedCalc-Pro Agent FrameworkModel=Llama-3.3-70B-Instruct2026.07 | 55.03 | 58.66 | 67.21 | 71.67 | 80.33 | |
| ReActModel=DeepSeek-V3.12026.07 | 53.23 | 51.61 | 38.11 | 20.97 | 24.19 | |
| MeNTiModel=MedResearcher-R1-32B2026.07 | 42.86 | 11.29 | 40.15 | 35.43 | 19.35 | |
| MeNTiModel=Qwen3-235B-A22B2026.07 | 37.96 | 54.84 | 59.53 | 45.16 | 58.06 | |
| MeNTiModel=gpt-oss-120b2026.07 | 36.52 | 56.45 | 61.68 | 61.29 | 54.83 | |
| MeNTiModel=DeepSeek-V3.12026.07 | 34.19 | 48.39 | 53.28 | 51.61 | 45.16 | |
| MeNTiModel=GPT-5-mini2026.07 | 31.78 | 66.13 | 56.08 | 54.83 | 53.22 | |
| MeNTiModel=Llama-3.3-70B-Instruct2026.07 | 21.12 | 33.9 | 15.5 | 16.12 | 11.29 | |
| MedRaCModel=gpt-oss-120b2026.07 | 0 | 0 | 0 | 0 | 0 | |
| MedRaCModel=Qwen3-235B-A22B2026.07 | 0 | 0 | 0 | 0 | 0 | |
| MedRaCModel=Llama-3.3-70B-Instruct2026.07 | 0 | 0 | 0 | 0 | 0 | |
| MedRaCModel=GPT-5-mini2026.07 | 0 | 0 | 0 | 0 | 0 | |
| MedRaCModel=DeepSeek-V3.12026.07 | 0 | 0 | 0 | 0 | 0 | |
| MedRaCModel=MedResearcher-R1-32B2026.07 | 0 | 0 | 0 | 0 | 0 |