List your company

Information Retrieval on HotpotQA (Trajectory and Final Answer Recall)

89.3RecallOpus-4.6
31.99646.87361.7576.627Jun 1, 2026

Evaluation results

MethodLinks
Opus-4.6Category=Frontier models2026.0689.389.391
GPT-5.4Category=Frontier models2026.06878789.8
Sonnet-4.6Category=Frontier models2026.0686.386.387.8
Harness-1Parameters=20B, Category=Open-source small models2026.0684.184.186
Kimi-K2.5Category=Frontier models2026.0677.677.683.7
Tongyi DeepResearchParameters=30B, Category=Open-source small models2026.0675.175.185.2
Context-1Parameters=20B, Category=Open-source small models2026.0674.674.683.5
GPT-OSSParameters=120B, Category=Frontier models2026.0665.765.789.7
Search-R1Parameters=32B, Category=Open-source small models2026.0645.445.445.4
GPT-OSSParameters=20B, Category=Open-source small models2026.0639.539.581.9
Qwen3Parameters=32B, Category=Open-source small models2026.0634.234.253.5