ManuSearch: Democratizing Deep Search in Large Language Models with a Transparent and Open Multi-Agent Framework

TL;DR

ManuSearch tackles the gap between closed, opaque deep-search systems and open, reproducible research by deploying a transparent, modular three-agent framework—solution planning, Internet search, and structured webpage reading—to perform deep, web-integrated reasoning. It introduces ORION, a challenging long-tail, open-web reasoning benchmark in English and Chinese, to rigorously evaluate such systems. Empirical results show ManuSearch substantially improves open-source baselines and competes with closed-source solutions, demonstrating the value of modular planning and tool-augmented retrieval for open deep search. The work provides a reproducible foundation (data and code) to advance research in open, trustworthy, deep-web reasoning frameworks.

Abstract

Recent advances in web-augmented large language models (LLMs) have exhibited strong performance in complex reasoning tasks, yet these capabilities are mostly locked in proprietary systems with opaque architectures. In this work, we propose \textbf{ManuSearch}, a transparent and modular multi-agent framework designed to democratize deep search for LLMs. ManuSearch decomposes the search and reasoning process into three collaborative agents: (1) a solution planning agent that iteratively formulates sub-queries, (2) an Internet search agent that retrieves relevant documents via real-time web search, and (3) a structured webpage reading agent that extracts key evidence from raw web content. To rigorously evaluate deep reasoning abilities, we introduce \textbf{ORION}, a challenging benchmark focused on open-web reasoning over long-tail entities, covering both English and Chinese. Experimental results show that ManuSearch substantially outperforms prior open-source baselines and even surpasses leading closed-source systems. Our work paves the way for reproducible, extensible research in open deep search systems. We release the data and code in https://github.com/RUCAIBox/ManuSearch

Figures (3)

• Figure 1: Distribution of domains in ORION.
• Figure 2: The overall framework of our proposed ManuSearch.
• Figure 3: A case of ManuSearch from GAIA benchmark.