List your company

EHR-based clinical information retrieval and question answering on AgentEHR-Bench (test)

58.5Diagnoses ScoreClaude Opus 4.6
16.27627.23838.249.162May 19, 2026

Evaluation results

MethodLinks
Claude Opus 4.6Model Access=Closed-source2026.0558.542.127.231.120.936
ClinSeek-35B-A3BModel Access=Open-source, Configuration=SFT2026.0555.438.527.631.716.734
Claude Sonnet 4.6Model Access=Closed-source2026.0554.435.623.426.323.732.7
MiniMax-M2.5Model Access=Open-source2026.0551.52919221727.7
Kimi K2.5Model Access=Open-source2026.0546.933.718.927.922.129.9
GLM-4.7Model Access=Open-source2026.0546.428.616.623.722.927.6
Qwen3.5-35B-A3B (base)Model Access=Open-source, Configuration=base2026.0536.617.716.221.918.122.1
Qwen3-235B-A22BModel Access=Open-source2026.0530.620.317.324.99.620.5
gpt-oss-120bModel Access=Open-source2026.0527.312.812.419.17.615.8
Tongyi DeepResearch 30B-A3BModel Access=Open-source2026.0525.814.98.817.913.216.1
OpenSeeker-30BModel Access=Open-source2026.0520.44.512.814.210.612.5
Gemma-4-26B-A4B-itModel Access=Open-source2026.0517.918.519.711.28.815.2