This setting trims Claude Code down for code-focused work: it lowers the output token limit and disables model calls that aren’t essential.
It also mutes cost warnings, keeping the workflow leaner.
The payoff is lower API spend and faster responses across long development sessions.
When to use
- On projects where you value code quality over conversational replies.
- When API cost is adding up and you want tighter control over tokens.
- To speed up repetitive programming tasks.
How to use
Apply the setting; if you later need longer answers, raise the output token limit.