Documentation quality evaluation on ClinicalBench 1.0 (test)
72.78IDEA ScoreAegle
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| AegleMethod Category=Ours2026.04 | 72.78 | 63.02 | 77.55 | 25.83 | |
| MDAgentsMethod Category=Medical multi-agent system2026.04 | 68.41 | 56.45 | 78.81 | 24.83 | |
| ToTMethod Category=Reasoning-strategy baseline2026.04 | 66.53 | 62.34 | 78.46 | 25.96 | |
| CoTMethod Category=Reasoning-strategy baseline2026.04 | 64.72 | 60.34 | 76.94 | 25.83 | |
| Qwen3-MaxProprietary=true2026.04 | 61.75 | 53.4 | 74.99 | 17.73 | |
| MedAgentsMethod Category=Medical multi-agent system2026.04 | 59.85 | 54.48 | 72 | 23.66 | |
| MiniMax-M2Open-source=true2026.04 | 57.78 | 46.18 | 73.87 | 16.41 | |
| Kimi K2 ThinkingOpen-source=true2026.04 | 54.05 | 42.86 | 70.76 | 17.1 | |
| Doubao-Seed-1.6Proprietary=true2026.04 | 51.51 | 39.8 | 69.03 | 18.04 | |
| DeepSeek-V3.2Open-source=true2026.04 | 50.51 | 38.64 | 71.73 | 17.33 | |
| Lingshu-32BOpen-source=true, Medical-domain=true2026.04 | 50.5 | 39.76 | 71.11 | 11.53 | |
| Gemini 2.5Proprietary=true2026.04 | 48.35 | 35.58 | 70.1 | 17.82 | |
| ERNIE-5.0-PreviewProprietary=true2026.04 | 47.09 | 35.55 | 67.19 | 15.82 | |
| GLM-4.6Open-source=true2026.04 | 47.02 | 34.7 | 68.48 | 16.73 | |
| DeepSeek-V3.2-ThinkingOpen-source=true2026.04 | 46.73 | 34.62 | 69.53 | 16.17 | |
| HuatuoGPT-o1-7BOpen-source=true, Medical-domain=true2026.04 | 46.54 | 37.98 | 68.66 | 9.62 | |
| Qwen3-8B-thinkOpen-source=true2026.04 | 43.18 | 27.66 | 66.74 | 10.19 | |
| GPT-4oProprietary=true2026.04 | 41.05 | 29.38 | 67.66 | 10.83 | |
| Baichuan-M2-32BOpen-source=true, Medical-domain=true2026.04 | 38.02 | 25.58 | 64 | 11.46 |