List your company

Domain Deep Research Tool Use on FinSearchComp Global-T2

70.6Success RateGemini-3-Pro
26.9238.2649.660.94Mar 10, 2026

Evaluation results

MethodLinks
Gemini-3-ProCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.0370.6
DIVE-8B (RL)Category=Ours, Temperature=1, Top-p=1, Number of runs=42026.0367.3
DIVE-8B (SFT)Category=Ours, Temperature=1, Top-p=1, Number of runs=42026.0362.1
DeepSeek-V3.2-ExpCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.0361.3
GPT-OSS-120BCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.0361
Claude-4-SonnetCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.0360.2
Kimi-K2-0905Category=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.0347.1
Gemini-2.5-ProCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.0344.5
EnvScaler-8BCategory=8B Baselines, Temperature=1, Top-p=1, Number of runs=42026.0340.7
WebExplorer-8BCategory=8B Baselines, Temperature=1, Top-p=1, Number of runs=42026.0335.9
SWE-Dev-8BCategory=8B Baselines, Temperature=1, Top-p=1, Number of runs=42026.0330.5
Qwen3-8B (base)Category=Ours, Temperature=1, Top-p=1, Number of runs=42026.0328.6