List your company

Data Analysis on DABStep hard

37.04AccuracyClaude Sonnet 4.5
0.993610.351819.7129.0682Jan 22, 2026

Evaluation results

MethodLinks
Claude Sonnet 4.5Model Type=Proprietary2026.0137.04
Claude Sonnet 4Model Type=Proprietary2026.0131.75
Kimi K2 InstructModel Type=Open-sourced2026.0128.84
GPT-5Reasoning effort=medium, Model Type=Proprietary2026.0128.31
Deepseek-v3.1Model Type=Open-sourced2026.0121.96
Qwen3 235B InstructModel Type=Open-sourced2026.0117.46
Qwen3-Coder 480BModel Type=Open-sourced2026.0114.29
GPT-5.1Reasoning effort=high, Model Type=Proprietary2026.0113.23
GPT-5.1Reasoning effort=none, Model Type=Proprietary2026.0111.9
GPT-OSS-120BModel Type=Open-sourced2026.017.94
GPT-4oModel Type=Proprietary2026.017.41
Qwen3-4B-InstructModel Type=Open-sourced2026.012.9
Qwen2.5-7B-InstructModel Type=Open-sourced2026.012.38