List your company

Evidence Retrieval on CORTEXBENCH Bridge

38.4Hit@5Gemini 3.1 Pro
19.57624.46329.3534.237Jun 29, 2026

Evaluation results

MethodLinks
Gemini 3.1 Pro2026.0638.414.1
GPT-4o2026.0633.427.9
GPT-5.42026.0624.428.6
DeepSeek-V4-Flash2026.0623.131.6
DeepSeek-V3.22026.0622.731.4
Gemini 3 Pro2026.0622.555
Claude Opus 4.62026.0620.640.8
Claude Sonnet 4.62026.0620.341.1