List your company

Safe Code Generation on Python Safe Code Generation Dataset

100Safe Code RateClaude Haiku 4.5
1.855227.335152.81578.2949Jun 15, 2026

Evaluation results

MethodLinks
Claude Haiku 4.5Prompting Condition=+Comp. I (SPARK Component I)2026.06100
Claude Sonnet 4.6Prompting Condition=+Comp. I (SPARK Component I)2026.0698.59
GPT-5.4Prompting Condition=+Comp. I (SPARK Component I)2026.0695.77
GPT-5.5Prompting Condition=+Comp. I (SPARK Component I)2026.0694.96
Claude Haiku 4.5Prompting Condition=Baseline2026.0691.55
Claude Sonnet 4.6Prompting Condition=Baseline2026.0691.55
GPT-5.4Prompting Condition=Baseline2026.0683.8
GPT-5.4-miniPrompting Condition=+Comp. I (SPARK Component I)2026.0681.69
DeepSeek V4 ProPrompting Condition=+Comp. I (SPARK Component I)2026.0680.28
GPT-5.5Prompting Condition=Baseline2026.0680.14
DeepSeek V4 FlashPrompting Condition=+Comp. I (SPARK Component I)2026.0671.83
GPT-5.4-miniPrompting Condition=Baseline2026.0662.68
DeepSeek V4 ProPrompting Condition=Baseline2026.0650
Claude Haiku 4.5Prompting Condition=Plain2026.0634.51
GPT-5.4Prompting Condition=Plain2026.0630.28
GPT-5.5Prompting Condition=Plain2026.0626.62
DeepSeek V4 FlashPrompting Condition=Baseline2026.0625.35
Claude Sonnet 4.6Prompting Condition=Plain2026.0611.97
DeepSeek V4 ProPrompting Condition=Plain2026.067.75
GPT-5.4-miniPrompting Condition=Plain2026.067.04
DeepSeek V4 FlashPrompting Condition=Plain2026.065.63