Financial Specialist Tool Use on Finance Agent Benchmark
39Success RateGemini-3-Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-3-ProCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.03 | 39 | |
| Claude-4-SonnetCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.03 | 39 | |
| GPT-OSS-120BCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.03 | 34 | |
| DIVE-8B (RL)Category=Ours, Temperature=1, Top-p=1, Number of runs=42026.03 | 34 | |
| Kimi-K2-0905Category=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.03 | 28 | |
| DIVE-8B (SFT)Category=Ours, Temperature=1, Top-p=1, Number of runs=42026.03 | 28 | |
| DeepSeek-V3.2-ExpCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.03 | 26 | |
| Gemini-2.5-ProCategory=Frontier (≫8B), Temperature=1, Top-p=1, Number of runs=42026.03 | 24 | |
| EnvScaler-8BCategory=8B Baselines, Temperature=1, Top-p=1, Number of runs=42026.03 | 14 | |
| WebExplorer-8BCategory=8B Baselines, Temperature=1, Top-p=1, Number of runs=42026.03 | 4 | |
| SWE-Dev-8BCategory=8B Baselines, Temperature=1, Top-p=1, Number of runs=42026.03 | 3 | |
| Qwen3-8B (base)Category=Ours, Temperature=1, Top-p=1, Number of runs=42026.03 | 2 |