List your company

Clinical Task Execution on MedAgentBench v2 (val)

77AccuracyGRASP
15.43231.41647.463.384May 28, 2026

Evaluation results

MethodLinks
GRASPModel=gpt-oss-120b2026.0577
GRASPModel=GPT-4.12026.0573.8
ExpeLModel=GPT-4.12026.0572.5
SkillXModel=GPT-4.12026.0572.1
GRASPModel=GPT-5.4 (low)2026.0571.2
ExpeLModel=GPT-5.4 (low)2026.0570.4
GRASPModel=Gemini 3.1 Flash Lite2026.0568.8
Seq. MemoryModel=GPT-4.12026.0568.8
Evo-MedAgentModel=GPT-5.4 (low)2026.0568.8
SkillXModel=GPT-5.4 (low)2026.0568.8
Batch MemoryModel=GPT-4.12026.0568.3
SkillXModel=gpt-oss-120b2026.0567.5
Evo-MedAgentModel=GPT-4.12026.0566.7
No skillsModel=GPT-4.12026.0565.9
ExpeLModel=gpt-oss-120b2026.0565.5
SkillXModel=Gemini 3.1 Flash Lite2026.0565.4
Evo-MedAgentModel=gpt-oss-120b2026.0565.3
No skillsModel=GPT-5.4 (low)2026.0563.5
Evo-MedAgentModel=Gemini 3.1 Flash Lite2026.0562.9
Batch MemoryModel=GPT-5.4 (low)2026.0562.9
Batch MemoryModel=gpt-oss-120b2026.0562.2
Seq. MemoryModel=GPT-5.4 (low)2026.0562.1
Seq. MemoryModel=Gemini 3.1 Flash Lite2026.0560
ExpeLModel=Gemini 3.1 Flash Lite2026.0558.3
No skillsModel=gpt-oss-120b2026.0557.6
Batch MemoryModel=Gemini 3.1 Flash Lite2026.0555.4
Seq. MemoryModel=gpt-oss-120b2026.0554.2
No skillsModel=Gemini 3.1 Flash Lite2026.0547.6
GRASPModel=DeepSeek V4 Flash2026.0536.8
ExpeLModel=DeepSeek V4 Flash2026.0531.5
Evo-MedAgentModel=DeepSeek V4 Flash2026.0525.5
No skillsModel=DeepSeek V4 Flash2026.0523.8
SkillXModel=DeepSeek V4 Flash2026.0519.5
Seq. MemoryModel=DeepSeek V4 Flash2026.0517.8
Batch MemoryModel=DeepSeek V4 Flash2026.0517.8