List your company

Policy Question Answering on PolicyBench

64.34AccuracyDeepseek-R1
57.964859.619961.27562.9301Apr 14, 2026

Evaluation results

MethodLinks
Deepseek-R12026.0464.34
Claude-3.7-sonnet2026.0464.13
Gemini-2.5-Flash2026.0463.82
Claude-3.5-Sonnet2026.0463.75
o4-mini2026.0462.98
QwQ-32B2026.0461.67
Gemini-2.0-Flash2026.0460.1
GPT-4o2026.0459.47
LLaMA-42026.0459.17
Deepseek-V32026.0459.1
Gemma 3-27B2026.0458.21