Broad Information Seeking on WideSearch
80.12Item F1 (Avg@4)Web2BigTable
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Web2BigTableBackbone LLM=GPT-5 mini + Gemini 3 Flash2026.04 | 80.12 | 82.48 | 63.53 | 65.12 | 38.5 | — | — | — | — | — | 40 | |
| Claude-4.5-SonnetSystem Architecture=Single Agent2026.04 | 65.7 | — | — | — | — | — | — | — | — | — | — | |
| Seed1.8System Architecture=Single Agent2026.04 | 63.8 | — | — | — | — | — | — | — | — | — | — | |
| GPT-5 HighSystem Architecture=Single Agent2026.04 | 62.2 | — | — | — | — | — | — | — | — | — | — | |
| Claude Sonnet 4 (Thinking)System Architecture=Multi-Agent Framework, Thinking Mode=true2026.04 | 62.2 | 73.1 | 38.5 | 52.2 | 3.6 | — | — | — | — | — | 6.5 | |
| Kimi K2System Architecture=Multi-Agent Framework2026.04 | 61.2 | 70.7 | 36.2 | 49.6 | 3 | — | — | — | — | — | 6.5 | |
| GeminiSystem Architecture=End-to-End System2026.04 | 59.1 | 67.2 | 36.6 | 45.4 | 4.3 | — | — | — | — | — | 8 | |
| Claude Sonnet 4 (Thinking)System Architecture=Single Agent, Thinking Mode=true2026.04 | 57.9 | 66.7 | 31.7 | 41.9 | 2.3 | — | — | — | — | — | 5 | |
| Gemini 2.5 ProSystem Architecture=Multi-Agent Framework2026.04 | 57.4 | 66.3 | 33.5 | 44.6 | 2 | — | — | — | — | — | 6.5 | |
| OpenAI o3-highSystem Architecture=Multi-Agent Framework2026.04 | 57.3 | 68.9 | 37.8 | 50.5 | 5.1 | — | — | — | — | — | 9.5 | |
| Gemini-3-ProSystem Architecture=Single Agent2026.04 | 57 | — | — | — | — | — | — | — | — | — | — | |
| Doubao-1.6System Architecture=Multi-Agent Framework2026.04 | 54.6 | 69.7 | 34 | 48.9 | 2.5 | — | — | — | — | — | 5.5 | |
| Kimi K2System Architecture=Single Agent2026.04 | 54.4 | 65.1 | 29.7 | 41.4 | 1.1 | — | — | — | — | — | 3.5 | |
| Doubao-1.6-non-thinkingSystem Architecture=Multi-Agent Framework, Thinking Mode=false2026.04 | 52.8 | 65.1 | 29.7 | 42.7 | 2.1 | — | — | — | — | — | 4.5 | |
| OpenAI o3-highSystem Architecture=Single Agent2026.04 | 52.6 | 62.3 | 34 | 44.1 | 4.5 | — | — | — | — | — | 9 | |
| Gemini 2.5 ProSystem Architecture=Single Agent2026.04 | 51 | 63.6 | 30 | 41.4 | 1.5 | — | — | — | — | — | 5 | |
| Doubao-1.6-non-thinkingSystem Architecture=Single Agent, Thinking Mode=false2026.04 | 49 | 62 | 27.2 | 39.9 | 1 | — | — | — | — | — | 3.5 | |
| ClaudeSystem Architecture=End-to-End System2026.04 | 48.4 | 58.5 | 24.1 | 33.5 | 2.5 | — | — | — | — | — | 5 | |
| Doubao-1.6System Architecture=Single Agent2026.04 | 48.3 | 63.9 | 30 | 44.1 | 2.6 | — | — | — | — | — | 5 | |
| OpenAI o3System Architecture=End-to-End System2026.04 | 45.5 | 56.5 | 23.9 | 36 | 3 | — | — | — | — | — | 5.5 | |
| DeepSeek-R1System Architecture=Multi-Agent Framework2026.04 | 44.3 | 60.3 | 22.9 | 36.6 | 0.8 | — | — | — | — | — | 3 | |
| DeepSeek-R1-671BSetting=Single Agent, Model=DeepSeek-R1-671B2026.02 | 41.3 | 55.1 | 20.7 | 31.7 | 40 | 150 | — | — | — | — | — | |
| DeepSeek-R1System Architecture=Single Agent2026.04 | 41.3 | 55.1 | 20.7 | 31.7 | 0.4 | — | — | — | — | — | 1.5 | |
| WIDESEEK-R1-4BSetting=Multi-Agent System, Model=WIDESEEK-R1-4B2026.02 | 40 | 51.8 | 15.3 | 24.4 | 40 | 100 | — | — | — | — | — | |
| GPT-5 miniSystem Architecture=Single Agent2026.04 | 33.28 | 44.1 | 22.1 | 32.4 | 4 | — | — | — | — | — | 7.5 | |
| Gemini 3 FlashSystem Architecture=Single Agent2026.04 | 31.61 | 42.3 | 18.7 | 28.5 | 3 | — | — | — | — | — | 6 | |
| Qwen3-4BSetting=Multi-Agent System, Model=Qwen3-4B2026.02 | 31.2 | 42.3 | 8.4 | 15.5 | 0 | 0 | — | — | — | — | — | |
| AgentFlow-7BSetting=Multi-Agent System, Model=AgentFlow-7B2026.02 | 28.7 | 45.4 | 9 | 20.2 | 40 | 150 | — | — | — | — | — | |
| SingleSeek-R1-4BSetting=Single Agent, Model=SingleSeek-R1-4B2026.02 | 28.1 | 39.2 | 6.5 | 12.5 | 30 | 100 | — | — | — | — | — | |
| MiroFlow-8BSetting=Multi-Agent System, Model=MiroFlow-8B2026.02 | 23.7 | 37.7 | 5.8 | 12.7 | 40 | 100 | — | — | — | — | — | |
| OWL-8BSetting=Multi-Agent System, Model=OWL-8B2026.02 | 20.2 | 29.3 | 3.1 | 5.8 | 0 | 0 | — | — | — | — | — | |
| Qwen3-4BSetting=Single Agent, Model=Qwen3-4B2026.02 | 20.1 | 30.2 | 3 | 4.8 | 0 | 0 | — | — | — | — | — | |
| ASearcher-7BSetting=Single Agent, Model=ASearcher-7B2026.02 | 16.5 | 26 | 2.8 | 5.8 | 0 | 0 | — | — | — | — | — | |
| Search-R1-7BSetting=Single Agent, Model=Search-R1-7B2026.02 | 15.5 | 24.4 | 2 | 4.4 | 0 | 0 | — | — | — | — | — | |
| Claude-S4 NoThinkReAct Type=SA2026.02 | — | — | — | — | — | — | 2.2 | 26.1 | 48.6 | 61.3 | — | |
| Claude-S4 NoThinkReAct Type=MA2026.02 | — | — | — | — | — | — | 3.2 | 33.7 | 56.6 | 68 | — | |
| Claude-S4 NoThink (TaS)ReAct Type=MA2026.02 | — | — | — | — | — | — | 3.5 | 36.7 | 60.5 | 74.7 | — | |
| Claude-S4 ThinkReAct Type=SA2026.02 | — | — | — | — | — | — | 2.3 | 31.7 | 57.9 | — | — | |
| Claude-S4 ThinkReAct Type=MA2026.02 | — | — | — | — | — | — | 3.6 | 38.5 | 62.2 | — | — | |
| Gemini-2.5-FlashReAct Type=SA2026.02 | — | — | — | — | — | — | 2 | 26.9 | 49.9 | 62.1 | — | |
| Gemini-2.5-FlashReAct Type=MA2026.02 | — | — | — | — | — | — | 1.9 | 26.3 | 45.7 | 55.4 | — | |
| Gemini-2.5-Flash (TaS)ReAct Type=MA2026.02 | — | — | — | — | — | — | 2.2 | 29.1 | 52.7 | 66.8 | — | |
| Gemini-2.5-ProReAct Type=SA2026.02 | — | — | — | — | — | — | 1.5 | 30 | 51 | — | — | |
| Gemini-2.5-ProReAct Type=MA2026.02 | — | — | — | — | — | — | 2 | 33.5 | 57.4 | — | — | |
| KIMI-K2ReAct Type=SA2026.02 | — | — | — | — | — | — | 1.1 | 29.7 | 54.4 | — | — | |
| KIMI-K2ReAct Type=MA2026.02 | — | — | — | — | — | — | 3 | 36.2 | 61.2 | — | — | |
| OpenAI o3ReAct Type=SA2026.02 | — | — | — | — | — | — | 4.5 | 34 | 52.6 | — | — | |
| OpenAI o3ReAct Type=MA2026.02 | — | — | — | — | — | — | 5.1 | 37.8 | 57.3 | — | — | |
| WebLeaperReAct Type=SA2026.02 | — | — | — | — | — | — | 4 | 31 | 48.8 | — | — |