PQS

PQS (Prompt Quality Score) is the world's first named AI prompt quality score, offered as an MCP server. It allows users to score, optimize, and compare LLM prompts before they are sent to any model, helping to prevent wasted inference spend on low-quality inputs. Built on frameworks like PEEM, RAGAS, G-Eval, and MT-Bench, PQS provides a comprehensive evaluation including a grade (A-F), a score out of 40, a percentile, and an 8-dimension breakdown. It supports various domain contexts for more accurate scoring and can be integrated as a pre-inference quality gate.

Key Features

Use Cases