List your company

Data Analysis Reasoning on DABStep (test)

90.12Accuracy (Easy)GPT-5-2
77.931281.095684.2687.4244Jun 4, 2026

Evaluation results

MethodLinks
GPT-5-2Framework=DataCOPE2026.0690.1256.4561.83
Claude Sonnet 4.5Framework=No Skill2026.0688.8927.3537.18
DeepSeek-V4-ProFramework=DataCOPE2026.0687.6653.5258.97
Claude Sonnet 4.5Framework=DataCOPE2026.0687.0457.462.13
Qwen3.5-397BFramework=DataCOPE2026.0687.0458.2262.82
GPT-5-2Framework=No Skill2026.0685.1916.7827.71
GPT-5-2Framework=Skill Creator2026.0685.1946.3652.56
DeepSeek-V4-ProFramework=Skill Creator2026.0682.139.9146.65
Claude Sonnet 4.5Framework=Skill Creator2026.0681.4847.8953.26
Claude Sonnet 4.6Framework=No Skill2026.0680.8633.5741.13
Qwen3.5-397BFramework=No Skill2026.0680.8629.8137.97
DeepSeek-V4-ProFramework=No Skill2026.0680.2511.9713.7
Qwen3.5-397BFramework=Skill Creator2026.0678.449.8854.44