WebSwarm: Recursive Multi-Agent Orchestration for Deep-and-Wide Web Search
2026-07-09 • Computation and Language
Computation and LanguageArtificial IntelligenceMultiagent Systems
AI summaryⓘ
The authors present WebSwarm, a system where many small search agents work together to tackle complex online searches that need both deep details and broad coverage. Unlike previous systems limited by single long searches or simple parallel efforts, WebSwarm lets agents break down tasks, delegate parts to others, and combine results step-by-step, guided by how information is structured on the web. Their experiments show WebSwarm does better than single or multi-agent methods on challenging search tasks, and they analyze why it works well. This framework improves cooperation and flexibility in web search agents.
large language modelsweb search agentsReAct frameworktask decompositionrecursive expansionmulti-agent systemsevidence-grounded searchagent collaborationBrowseComp-PlusDeepWideSearch
Authors
Xiaoshuai Song, Liancheng Zhang, Kangzhi Zhao, Yutao Zhu, Zhongyuan Wang, Guanting Dong, Jinghan Yang, Han Li, Kun Gai, Ji-Rong Wen, Zhicheng Dou
Abstract
Large language model (LLM)-based web search agents are transforming information seeking from simple factoid question answering into complex, deep-and-wide search and research-oriented tasks. A single ReAct-style agent is constrained by one long trajectory and limited context, making it difficult to handle depth and coverage simultaneously. Existing multi-agent systems improve search coverage through parallel execution and aggregation, but still exhibit clear limitations in recursive depth, collaboration adaptability, and evidence-grounded expansion. We propose WebSwarm, a progressive recursive delegation framework that jointly constructs task decomposition, recursive expansion, and agent collaboration during inference. WebSwarm dynamically instantiates agentic search nodes, each coupling a local objective with a search mode that specifies how the node should organize search and collaboration. Each node can either solve its objective itself or further delegate child nodes; after solving, it returns evidence and results upward, enabling parent nodes to further expand, revise, or aggregate the search process. To guide this process, WebSwarm first probes how task-relevant information is organized on the web to ground subsequent node expansion, and reuses process-level experience across homogeneous sibling nodes. Experiments on BrowseComp-Plus, WideSearch, DeepWideSearch, and GISA show that WebSwarm consistently outperforms single-agent and multi-agent baselines on deep, wide, and interleaved deep-and-wide tasks. Further analyses of ablation, task difficulty, web tool efficiency, and model generalization explain WebSwarm's effectiveness and provide insights for multi-agent search systems.