DEVOURED
The text in Claude Code's “Extended Thinking” output is not authentic
The full reasoning process behind Claude Code’s 'Extended Thinking' feature is encrypted and inaccessible to standard users.
What: The 'Extended Thinking' output provided in the API is a condensed summary rather than the original step-by-step reasoning, which Anthropic keeps encrypted on its servers. Full transparency requires specific enterprise-level agreements.
Original article
Claude Code's 'Extended Thinking' reasoning is encrypted. Anthropic holds the key, and users' machines never receive it. The API returns a summary of the reasoning, and not the reasoning itself. Receiving the full thinking output requires an enterprise agreement.
Explore all 42 articles in today's digest
Fresh dev news, AI-digested daily. Never miss a trend at devoured.fyi