Anthropic added a feature in Claude Code that lets the model use its judgment to choose a lower-power model for simple coding tasks, reducing token consumption and costs, crucial for MENA enterprises optimizing AI efficiency.

1 min read

Anthropic Launches Autonomous Judgment in Claude Code: Smart Cost Reduction

Overview

FAQ

What is the autonomous judgment feature in Claude Code?

It allows the model to use its judgment to select a lower-power model (e.g., Sonnet or Haiku) for simple coding tasks, while complex tasks remain with the main model.

How does this compare to traditional methods?

Traditional methods require manual model selection per task, while this feature automates the process, saving time and costs.

Should MENA teams adopt this now?

Yes, especially for organizations using Claude Code heavily, as it can reduce costs by up to 40% based on early reports.

Source: Simon Willison (LLM & tools)

AI-assisted content, human-reviewed.