AgentGroupChat-V2: Divide-and-Conquer Is What LLM-Based Multi-Agent System Need

TL;DR

AgentGroupChat-V2 introduces a divide-and-conquer, adaptive multi-agent framework that significantly improves performance and scalability in complex reasoning tasks across diverse domains.

Contribution

The paper presents a novel multi-agent system architecture with hierarchical task decomposition and adaptive collaboration, enhancing generalizability and efficiency over existing frameworks.

Findings

Achieves 91.50% accuracy on GSM8K, surpassing baselines by 5.6%

Nearly doubles accuracy on AIME compared to other methods

Attains 79.20% pass@1 on HumanEval, with notable improvements on complex tasks

Abstract

Large language model based multi-agent systems have demonstrated significant potential in social simulation and complex task resolution domains. However, current frameworks face critical challenges in system architecture design, cross-domain generalizability, and performance guarantees, particularly as task complexity and number of agents increases. We introduces AgentGroupChat-V2, a novel framework addressing these challenges through three core innovations: (1) a divide-and-conquer fully parallel architecture that decomposes user queries into hierarchical task forest structures enabling dependency management and distributed concurrent processing. (2) an adaptive collaboration engine that dynamically selects heterogeneous LLM combinations and interaction modes based on task characteristics. (3) agent organization optimization strategies combining divide-and-conquer approaches for efficient problem decomposition. Extensive experiments demonstrate AgentGroupChat-V2's superior performance across diverse domains, achieving 91.50% accuracy on GSM8K (exceeding the best baseline by 5.6 percentage points), 30.4% accuracy on competition-level AIME (nearly doubling other methods), and 79.20% pass@1 on HumanEval. Performance advantages become increasingly pronounced with higher task difficulty, particularly on Level 5 MATH problems where improvements exceed 11 percentage points compared to state-of-the-art baselines. These results confirm that AgentGroupChat-V2 provides a comprehensive solution for building efficient, general-purpose LLM multi-agent systems with significant advantages in complex reasoning scenarios. Code is available at https://github.com/MikeGu721/AgentGroupChat-V2.