[BUG] Unreasonable token usage for simple tasks in Claude Code using Sonnet or Opus

Resolved 💬 2 comments Opened Apr 10, 2026 by rhayann1337 Closed May 24, 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?

Claude Code is consuming an excessive amount of tokens even for simple, well-defined tasks.

A basic request to integrate two backend endpoints into an already structured frontend consumed 22% of a 5-hour quota just on /plan, without any implementation.

Additionally, on an Enterprise plan, we’ve seen ~$180 in usage over just 4 days, despite using the tool correctly and following best practices.

This level of consumption is not sustainable and makes the tool impractical for real-world development. It’s reaching a point where we are considering switching to alternative tools, even at the cost of some quality.

Token usage needs to be significantly more efficient and predictable, especially for low-complexity tasks.

What Should Happen?

Token consumption should be predictable and efficient, allowing consistent usage without rapidly exhausting quotas or generating unexpectedly high costs.

Error Messages/Logs

N/A

Steps to Reproduce

Use CLI with opus or sonnet

Claude Model

Not sure / Multiple models

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.98

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_

View original on GitHub ↗

This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗