[BUG] Credits consume at alariming rate : 5-8 times faster than normal

Resolved 💬 3 comments Opened Apr 3, 2026 by sympl-marketing Closed Apr 7, 2026

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?

since yesterday afternoon, my credits burn at an alarmin rate. My guess is that they burn 5-8, possilby 10 times faster than normal. I am at a max plan. I am used to running 3 sessions at the same time, working 4 of the 5 hours in Claude. In one month, I hit my 5-hr widow max once, while I'm almost working full time in Claude. Yesterday afternoon, I started noticing strange behaviour, eg. my max was reached after 4 hours, while I had not been really productive. Picked it up again after 18h --> after 80 minutes of monterminal work --> 84%. I then stopped working, because insane. Today, I did nothing in Claude except for the last 2 hours : 2 questions : a) how do I strikkethrough in an md document in green + why are my credits burning like mad. This is really the only things I did --> 30% (yes THIRTY ! ). This must be a bug. You also have to know that I have optimised my context window by organising claude.md and memory.md at 4 different levels (me, company, departement, sub-department. I have also instrcuted at top claude.md level to always do model switching depending on complecity of task. So I know that I have to carefully manage context and model/. But in a couple of days --> credit usage has gone +/- x 5-8

What Should Happen?

Much less credit consumption

Error Messages/Logs

Steps to Reproduce

cannot do this

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.91

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_

View original on GitHub ↗

This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗