CodeAgents: A Token-Efficient Framework for Codified Multi-Agent Reasoning in LLMs
This work addresses the need for scalable and token-efficient multi-agent systems in AI, offering a novel approach that is incremental but impactful for domain-specific applications like virtual environments and QA tasks.
The paper tackles the problem of inefficient and unstructured prompting for multi-agent reasoning in LLMs by introducing CodeAgents, a framework that codifies agent interactions into modular pseudocode, resulting in improved planning performance (e.g., 3-36 percentage point gains) and significant token reductions (55-87% for input, 41-70% for output).
Effective prompt design is essential for improving the planning capabilities of large language model (LLM)-driven agents. However, existing structured prompting strategies are typically limited to single-agent, plan-only settings, and often evaluate performance solely based on task accuracy - overlooking critical factors such as token efficiency, modularity, and scalability in multi-agent environments. To address these limitations, we introduce CodeAgents, a prompting framework that codifies multi-agent reasoning and enables structured, token-efficient planning in multi-agent systems. In CodeAgents, all components of agent interaction - Task, Plan, Feedback, system roles, and external tool invocations - are codified into modular pseudocode enriched with control structures (e.g., loops, conditionals), boolean logic, and typed variables. This design transforms loosely connected agent plans into cohesive, interpretable, and verifiable multi-agent reasoning programs. We evaluate the proposed framework across three diverse benchmarks - GAIA, HotpotQA, and VirtualHome - using a range of representative LLMs. Results show consistent improvements in planning performance, with absolute gains of 3-36 percentage points over natural language prompting baselines. On VirtualHome, our method achieves a new state-of-the-art success rate of 56%. In addition, our approach reduces input and output token usage by 55-87% and 41-70%, respectively, underscoring the importance of token-aware evaluation metrics in the development of scalable multi-agent LLM systems. The code and resources are available at: https://anonymous.4open.science/r/CodifyingAgent-5A86