List your company

Clinical Task Execution on MedAgentBench (test)

88.8AccuracyGRASP
32.22446.91261.676.288May 28, 2026

Evaluation results

MethodLinks
GRASPModel=gpt-oss-120b2026.0588.8
GRASPModel=GPT-5.4 (low)2026.0585.4
GRASPModel=GPT-4.12026.0584.9
GRASPModel=Gemini 3.1 Flash Lite2026.0571.4
GRASPModel=DeepSeek V4 Flash2026.0570
Evo-MedAgentModel=gpt-oss-120b2026.0567.8
SkillXModel=DeepSeek V4 Flash2026.0555.9
ExpeLModel=DeepSeek V4 Flash2026.0555.6
Evo-MedAgentModel=Gemini 3.1 Flash Lite2026.0555.2
No skillsModel=Gemini 3.1 Flash Lite2026.0554.2
Seq. MemoryModel=Gemini 3.1 Flash Lite2026.0554.2
SkillXModel=Gemini 3.1 Flash Lite2026.0554.2
ExpeLModel=Gemini 3.1 Flash Lite2026.0553.6
SkillXModel=gpt-oss-120b2026.0553.1
Evo-MedAgentModel=DeepSeek V4 Flash2026.0551.9
ExpeLModel=gpt-oss-120b2026.0549.1
No skillsModel=DeepSeek V4 Flash2026.0547.7
Batch MemoryModel=GPT-4.12026.0547.4
Seq. MemoryModel=GPT-5.4 (low)2026.0547.4
Batch MemoryModel=GPT-5.4 (low)2026.0546.4
No skillsModel=GPT-4.12026.0545.5
Seq. MemoryModel=GPT-4.12026.0545.3
ExpeLModel=GPT-4.12026.0545.3
Evo-MedAgentModel=GPT-4.12026.0545.3
No skillsModel=GPT-5.4 (low)2026.0545.1
SkillXModel=GPT-5.4 (low)2026.0544.8
ExpeLModel=GPT-5.4 (low)2026.0543.8
Evo-MedAgentModel=GPT-5.4 (low)2026.0543.8
Seq. MemoryModel=gpt-oss-120b2026.0541.2
No skillsModel=gpt-oss-120b2026.0540.6
Batch MemoryModel=Gemini 3.1 Flash Lite2026.0539.6
SkillXModel=GPT-4.12026.0538.5
Batch MemoryModel=DeepSeek V4 Flash2026.0537.8
Seq. MemoryModel=DeepSeek V4 Flash2026.0535
Batch MemoryModel=gpt-oss-120b2026.0534.4