AI Agent Protocol Benchmark: A Unified Framework for Evaluating Multi-Agent Communication
Published in arXiv preprint, 2025
This paper presents a unified benchmark for evaluating AI agent communication protocols across diverse tasks, focusing on task performance, communication cost, and robustness.
Recommended citation: Li, J., et al. (2025). "AI Agent Protocol Benchmark: A Unified Framework for Evaluating Multi-Agent Communication." arXiv preprint arXiv:2510.17149.
Download Paper
