Prompt-Optimisation

How tighter instructions change which models pass your quality bar

Five models. Two prompts. Same task, same context, same rubric. On the vague prompt every model scored exactly 62.5. On the specific prompt the scores split by 40 points.

The model did not change. The prompt did.

Vague prompt

Both runs -- all five models: 62.5

Zero variance. The prompt gave no format instruction. Every model produced multi-section output with headers and bullet points. Brevity: 0. Actionability: partial. Claude at $48.68/month and Nano at $0.26/month scored identically.

Sep 23, 2026