Wide Search on WideSearch 40 samples
9.5ReAct AccOpenAI o3
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| OpenAI o3Type=MA, Max@k=4, Framework=Foundation Models with Tools2026.02 | 9.5 | 50.5 | 68.9 | — | — | |
| Claude-S4 NoThink (TaS)Type=MA, Max@k=4, Framework=TaS Framework, Base Model=Claude-Sonnet-42026.02 | 9.1 | 49 | 71 | 84.4 | — | |
| OpenAI o3Type=SA, Max@k=4, Framework=Foundation Models with Tools2026.02 | 9 | 44.1 | 62.3 | — | — | |
| Claude-S4 ThinkType=MA, Max@k=4, Framework=Foundation Models with Tools, Base Model=Claude-Sonnet-42026.02 | 6.5 | 52.2 | 73.1 | — | — | |
| Gemini-2.5-ProType=MA, Max@k=4, Framework=Foundation Models with Tools2026.02 | 6.5 | 44.6 | 66.3 | — | — | |
| KIMI-K2Type=MA, Max@k=4, Framework=Foundation Models with Tools2026.02 | 6.5 | 49.6 | 70.7 | — | — | |
| Claude-S4 ThinkType=SA, Max@k=4, Framework=Foundation Models with Tools, Base Model=Claude-Sonnet-42026.02 | 5 | 41.9 | 66.7 | — | — | |
| Gemini-2.5-ProType=SA, Max@k=4, Framework=Foundation Models with Tools2026.02 | 5 | 41.4 | 63.6 | — | — | |
| Gemini-2.5-FlashType=SA, Max@k=4, Framework=TaS Framework Context2026.02 | 5 | 41.1 | 64.8 | 78 | — | |
| Gemini-2.5-Flash (TaS)Type=MA, Max@k=4, Framework=TaS Framework2026.02 | 5 | 45.7 | 67.6 | 82.2 | — | |
| Gemini-2.5-FlashType=MA, Max@k=4, Framework=TaS Framework Context2026.02 | 4.5 | 42.3 | 61.7 | 71.4 | — | |
| Claude-S4 NoThinkType=SA, Max@k=4, Framework=TaS Framework Context, Base Model=Claude-Sonnet-42026.02 | 4.5 | 38.1 | 60.9 | 74.1 | — | |
| Claude-S4 NoThinkType=MA, Max@k=4, Framework=TaS Framework Context, Base Model=Claude-Sonnet-42026.02 | 4 | 46.8 | 66.9 | 78.2 | — | |
| KIMI-K2Type=SA, Max@k=4, Framework=Foundation Models with Tools2026.02 | 3.5 | 41.4 | 65.1 | — | — |