MCPAgentBench: A Real-world Task Benchmark for Evaluating LLM Agent MCP Tool Use

TL;DR

MCPAgentBench is a new benchmark designed to evaluate the tool-use capabilities of LLM-based agents in real-world scenarios, addressing current evaluation limitations with authentic tasks, simulated MCP tools, and dynamic testing environments.

Contribution

It introduces MCPAgentBench, a comprehensive benchmark with real-world MCP tasks, simulated tools, and metrics to assess LLM agents' tool selection and execution efficiency.

Findings

Significant performance differences among LLMs in complex tasks

Effective evaluation of tool discrimination abilities

Benchmark facilitates future LLM agent development

Abstract

Large Language Models (LLMs) are increasingly serving as autonomous agents, and their utilization of external tools via the Model Context Protocol (MCP) is considered a future trend. Current MCP evaluation sets suffer from issues such as reliance on external MCP services and a lack of difficulty awareness. To address these limitations, we propose MCPAgentBench, a benchmark based on real-world MCP definitions designed to evaluate the tool-use capabilities of agents. We construct a dataset containing authentic tasks and simulated MCP tools. The evaluation employs a dynamic sandbox environment that presents agents with candidate tool lists containing distractors, thereby testing their tool selection and discrimination abilities. Furthermore, we introduce comprehensive metrics to measure both task completion rates and execution efficiency. Experiments conducted on various latest mainstream Large Language Models reveal significant performance differences in handling complex, multi-step tool invocations. All code is open-source at Github.