List your company

Evidence Retrieval on CORTEXBENCH Comparison

46Hit@5Gemini 3.1 Pro
23.74429.52235.341.078Jun 29, 2026

Evaluation results

MethodLinks
Gemini 3.1 Pro2026.064612.6
GPT-4o2026.0636.633.9
DeepSeek-V4-Flash2026.0630.545.5
Claude Sonnet 4.62026.0629.835.7
GPT-5.42026.0629.639.5
Claude Opus 4.62026.0629.149.2
DeepSeek-V3.22026.0628.645.2
Gemini 3 Pro2026.0624.650.9