Deep Research on DeepSearchQA
80ScoreClaude-4.5-Opus
Evaluation results
| Method | Links | |
|---|---|---|
| Claude-4.5-OpusModel Category=Foundation Models2026.04 | 80 | |
| GPT-5 HighModel Category=Foundation Models2026.04 | 79 | |
| Kimi-K2.5Model Category=Foundation Models2026.04 | 77.1 | |
| Gemini-3-ProModel Category=Foundation Models2026.04 | 76.9 | |
| DeepSeek-V3.2Model Category=Foundation Models2026.04 | 60.9 | |
| DR-Venus-4B-RLModel Category=Trained Agents, Model Scale Group=≤9B, Training Protocol=RL2026.04 | 39.6 | |
| DR-Venus-4B-SFTModel Category=Trained Agents, Model Scale Group=≤9B, Training Protocol=SFT2026.04 | 37.7 | |
| AgentCPM-Explore-4BModel Category=Trained Agents, Model Scale Group=≤9B2026.04 | 32.8 | |
| WebExplorer-8B-RLModel Category=Trained Agents, Model Scale Group=≤9B, Training Protocol=RL2026.04 | 17.8 |