A new trend is sweeping through the AI community: reducing instructions and responses to the essentials to cut back on token usage.

One of the most popular examples is GitHub's Caveman plugin, which makes models like Codex, Claude, and Gemini communicate in ultra-minimal phrases like, "Claude thinks. Claude writes. Claude is ready."

The result? Token usage can be reduced by 65-75%, decreasing costs while keeping the same task ongoing.

Their motto says it all: "Save token. Save money."