List your company

Long-Horizon Stability on BioProBench Subset C (test)

100Success RateBioProAgent
30.63248.64166.6584.659Mar 1, 2026

Evaluation results

MethodLinks
BioProAgentBackbone=Gemini-3-Flash2026.0310071.895
ReActBackbone=Gemini-3-Flash2026.0388.911.421.7
AutoGPTBackbone=Gemini-3-Flash2026.0366.740.964.4
ReflexionBackbone=Gemini-3-Flash2026.0333.300