List your company

Clinical Task Execution on MedAgentBench OOD

80.6AccuracyGRASP
-3.22418.53840.362.062May 28, 2026

Evaluation results

MethodLinks
GRASPModel=GPT-5.4 (low)2026.0580.6
GRASPModel=gpt-oss-120b2026.0556.3
GRASPModel=DeepSeek V4 Flash2026.0552.3
GRASPModel=Gemini 3.1 Flash Lite2026.0541.7
SkillXModel=Gemini 3.1 Flash Lite2026.0538.3
Evo-MedAgentModel=gpt-oss-120b2026.0531.3
GRASPModel=GPT-4.12026.0527.2
SkillXModel=gpt-oss-120b2026.0521.3
Evo-MedAgentModel=DeepSeek V4 Flash2026.0520
No skillsModel=Gemini 3.1 Flash Lite2026.0518.1
ExpeLModel=Gemini 3.1 Flash Lite2026.0517.2
Seq. MemoryModel=gpt-oss-120b2026.0516.7
Batch MemoryModel=GPT-5.4 (low)2026.0516.7
Evo-MedAgentModel=Gemini 3.1 Flash Lite2026.0515.6
Seq. MemoryModel=GPT-5.4 (low)2026.0515.6
Batch MemoryModel=gpt-oss-120b2026.0513
ExpeLModel=gpt-oss-120b2026.0512
ExpeLModel=GPT-5.4 (low)2026.0511.7
Seq. MemoryModel=Gemini 3.1 Flash Lite2026.058.9
No skillsModel=gpt-oss-120b2026.058.7
Seq. MemoryModel=DeepSeek V4 Flash2026.058.7
Batch MemoryModel=Gemini 3.1 Flash Lite2026.058.3
SkillXModel=DeepSeek V4 Flash2026.056.3
Batch MemoryModel=DeepSeek V4 Flash2026.053.3
SkillXModel=GPT-5.4 (low)2026.053.3
No skillsModel=DeepSeek V4 Flash2026.053.1
ExpeLModel=DeepSeek V4 Flash2026.052.3
Batch MemoryModel=GPT-4.12026.052.2
No skillsModel=GPT-5.4 (low)2026.051.7
Evo-MedAgentModel=GPT-5.4 (low)2026.051.7
Seq. MemoryModel=GPT-4.12026.050.6
ExpeLModel=GPT-4.12026.050.6
Evo-MedAgentModel=GPT-4.12026.050.6
No skillsModel=GPT-4.12026.050.3
SkillXModel=GPT-4.12026.050