Clinical Diagnostic Reasoning Benchmark 1.0 (test)
50.66ICD RecallDr. Assistant
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Dr. AssistantNote=Primary proposed model2026.01 | 50.66 | 47.17 | |
| Gemini-3-Pro-PreviewAccess=Closed-source2026.01 | 47.48 | 53.33 | |
| GPT-5Access=Closed-source2026.01 | 46.51 | 48.42 | |
| HuatuoGPT-o1-72BAccess=Open-source, Domain=Medical2026.01 | 44.6 | 52.11 | |
| Qwen3-32BAccess=Open-source, Domain=Non-medical2026.01 | 43.56 | 46.6 | |
| Qwen3-14BAccess=Open-source, Domain=Non-medical2026.01 | 42.98 | 46 | |
| Gemini-2.5-FlashAccess=Closed-source2026.01 | 42.88 | 44.97 | |
| Grok 4 FastAccess=Closed-source2026.01 | 42.84 | 51.59 | |
| DeepSeek-R1-Distill-14BAccess=Open-source, Domain=Non-medical2026.01 | 39.13 | 42.67 | |
| Doctor-R1-8BAccess=Open-source, Domain=Medical2026.01 | 38.73 | 48 | |
| Baichuan-M2-32BAccess=Open-source, Domain=Medical2026.01 | 36.87 | 45.17 | |
| Med42-v2-8BAccess=Open-source, Domain=Medical2026.01 | 27.87 | 30.81 | |
| Llama3-OpenBioLLM-70BAccess=Open-source, Domain=Medical2026.01 | 22.16 | 27.55 |