[BUG] Token consumption is too fast recently

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

I'm a Pro member. I asked 3-4 questions on the web chat, and it has already consumed 9% of my 5-hour quota. I'm a Pro member.

What Should Happen?

Previously, it was estimated to be only 1% to 2%. Token consumption has increased by 5-10 times.

Error Messages/Logs

Steps to Reproduce

1. Open https://claude.ai/
2. Log in with a Pro account
3. Ask a few questions about ISO/IEC/IEEE 29148:2018

• Introduce ISO/IEC/IEEE 29148:2018
• Generate an introduction to StRS (Stakeholder Requirements Specification).
• Stakeholder Requirements Specification，Generate an introduction to SyRS (System Requirements Specification).
• Following the same format as above, generate an introduction to SRS (Software Requirements Specification).

Claude Model

Sonnet (default)

Is this a regression?

I don't know

Last Working Version

_No response_

Claude Code Version

web chat

Platform

Other

Operating System

Other

Terminal/Shell

Other

Additional Information

_No response_