LLM API 호출 65% 낭비 제거를 위한 Local-first Prompt Router 설계
I Tracked Every LLM API Call For a Week — 65% Were Unnecessary
I Tracked Every LLM API Call For a Week — 65% Were Unnecessary
I built a VS Code extension that auto-selects Ask/Plan/Agent mode for every Copilot prompt
Growing void between enterprise and frontier AI puts open weights models in the spotlight