Medical Dialogue on MTMedDialog Sample 15 cases per department
69.32Overall AccuracyTheraAgent
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| TheraAgent2026.05 | 69.32 | 75.2 | 73.44 | 76.27 | 58 | |
| Claude-4-Sonnet2026.05 | 63.92 | 63.2 | 67.68 | 63.2 | 58.13 | |
| Gemini-2.5-Pro2026.05 | 61.63 | 66.4 | 65.45 | 64.93 | 52.79 | |
| DeepSeek-R12026.05 | 57.6 | 51.2 | 62 | 60.27 | 50 | |
| Kimi-K22026.05 | 55.72 | 55.2 | 59.76 | 55.2 | 49.33 |