vLLM Benchmark

MCP vLLM Benchmarking Tool enables interactive performance testing of vLLM deployments through a simple interface. This proof-of-concept implementation allows users to benchmark language models served by vLLM by specifying endpoints, model names, and test parameters through natural language prompts. The tool leverages code from vLLM's official benchmarking suite to measure metrics like throughput, latency, and token generation speed across multiple test iterations. Developed by Eliovp-BV as an exploration of MCP capabilities, it's useful for AI engineers who need to evaluate and compare the performance characteristics of different model deployments.