Agentic Search on Agentic-Wide XBench (test)
74.19AccuracyA-MAPREDUCE
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| A-MAPREDUCEBackbone=GPT-5-mini2026.02 | 74.19 | 707.32 | 0.35 | |
| Flash-SearcherBackbone=GPT-5-mini2026.02 | 67.74 | 679.3 | 0.25 | |
| SmolAgentsBackbone=GPT-5-mini2026.02 | 51.61 | 1,335.1 | 0.57 | |
| OWLBackbone=GPT-5-mini2026.02 | 45.16 | 1,003.27 | 0.13 |