List your company

Question Answering on SearchGymBench Complex QA 1.0 (test)

47.6AccuracyDeepSeek-V3.2
1.52813.48925.4537.411Jan 21, 2026

Evaluation results

MethodLinks
DeepSeek-V3.2Category=Commercial Model2026.0147.6
SearchGymBackbone=Qwen2.5-7B2026.0141.1
SearchGymBackbone=Qwen3-8B2026.0140.6
Kimi-k2Category=Commercial Model2026.0124.4
SearchGymBackbone=Qwen3-4B2026.0120
ASearcher-Web-QwQCategory=Open-Source Baseline2026.0117.8
ASearcher-Web-7BCategory=Open-Source Baseline2026.017.2
Qwen3-8BCategory=Open-Source Baseline2026.013.3
Qwen3-4BCategory=Open-Source Baseline2026.013.3