Flash-Searcher: Fast and Effective Web Agents via DAG-Based Parallel Execution
Authors:
- Tianrui QinTQTianrui Qin
- Qianben ChenQCQianben Chen
- Sinuo WangSWSinuo Wang
- He XingHXHe Xing
- King ZhuKZKing Zhu
- He ZhuHZHe Zhu
- Dingfeng ShiDSDingfeng Shi
- Xinxin LiuXLXinxin Liu
- Ge ZhangGZGe Zhang
- Jiaheng LiuJLJiaheng Liu
- Yuchen Eleanor JiangYJYuchen Eleanor Jiang
- Xitong GaoXGXitong Gao
Preprint
Abstract
Large language models (LLMs) have demonstrated remarkable capabilities in complex reasoning tasks when equipped with external tools. However, current frameworks predominantly rely on sequential processing, leading to inefficient execution particularly for tasks requiring extensive tool interaction. This paper introduces Flash-Searcher, a novel parallel agent reasoning framework that fundamentally reimagines the execution paradigm from sequential chains to directed acyclic graphs (DAGs). Flash-Searcher decomposes complex tasks into subtasks with explicit dependencies, enabling concurrent execution of independent reasoning paths while maintaining logical constraints. Through dynamic workflow optimization, our framework continuously refines the execution graph based on intermediate results, effectively integrating summary module. Comprehensive evaluations across multiple benchmarks demonstrate that Flash-Searcher consistently outperforms existing approaches. Specifically, it achieves 67.7% accuracy on BrowseComp and 83% on xbench-DeepSearch, while reducing agent execution steps by up to 35% compared to current frameworks. Furthermore, when distilling this parallel reasoning pipeline into single models, we observe substantial performance gains across diverse backbone architectures, underscoring the generalizability of our methodology. Our work thus represents a significant advance in agent architecture design, offering a more scalable and efficient paradigm for complex reasoning tasks.Keywords
All Related Versions
This publication has 11 references indexed in Scilit:
- Symbolic learning enables self-evolving agentsAI Open, 2025
- DeepResearcher: Scaling Deep Research via Reinforcement Learning in Real-world Environments2025
- EvolveSearch: An Iterative Self-Evolving Search Agent2025
- SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesis2025
- Search-o1: Agentic Search-Enhanced Large Reasoning Models2025
- OS Agents: A Survey on MLLM-based Agents for Computer, Phone and Browser Use2025
- WebWalker: Benchmarking LLMs in Web Traversal2025
- ChatDev: Communicative Agents for Software Development2024
- Tree of Thoughts: Deliberate Problem Solving with Large Language Models2023
- Multi-Agent Reinforcement Learning: A Review of Challenges and ApplicationsApplied Sciences, 2021