List your company

Code Generation on HumanEval (pass@1 and Time)

86.2pass@1GraphFlow
65.60870.95476.381.646May 21, 2026

Evaluation results

MethodLinks
GraphFlowModel=Qwen-2.5-7B2026.0586.215.76
GraphFlowModel=Gemma-2-9B2026.0582.516.65
AFlowModel=Qwen-2.5-7B2026.0578.118.55
AutoFlowModel=Qwen-2.5-7B2026.0577.418.3
AgentKBModel=Qwen-2.5-7B2026.0576.820.71
GraphFlowModel=Llama-3.1-8B2026.0576.613.45
AFlowModel=Gemma-2-9B2026.0575.419.5
TaskWeaverModel=Qwen-2.5-7B2026.0575.325.76
LLMCompilerModel=Qwen-2.5-7B2026.057412.29
AgentKBModel=Gemma-2-9B2026.0573.818.5
AutoFlowModel=Gemma-2-9B2026.0572.520.4
AFlowModel=Llama-3.1-8B2026.0572.217.1
MetaGPTModel=Qwen-2.5-7B2026.0572.112.2
AgentKBModel=Llama-3.1-8B2026.0571.218.05
LLMCompilerModel=Gemma-2-9B2026.0571.216.85
AutoFlowModel=Llama-3.1-8B2026.057119.55
MetaGPTModel=Gemma-2-9B2026.0570.715.69
LLMCompilerModel=Llama-3.1-8B2026.0570.514.52
TaskWeaverModel=Gemma-2-9B2026.0570.525.25
MetaGPTModel=Llama-3.1-8B2026.0570.213.62
TaskWeaverModel=Llama-3.1-8B2026.0569.924.8
VanillaModel=Qwen-2.5-7B2026.0569.27.23
VanillaModel=Gemma-2-9B2026.0566.813.78
VanillaModel=Llama-3.1-8B2026.0566.48.76