[BUG] Excessive token consumption on Max plan - possible tracking issue

Open 💬 0 comments Opened Jun 18, 2026 by thearchitechau

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?

/feedback Excessive token consumption on Max plan - possible tracking issue

I'm experiencing unusually high token consumption that doesn't match my usage pattern:

Consumption rate: ~1% of Max plan limit every 30-45 minutes

Current setup: 2-3 Sonnet agents, medium/high effort, thinking disabled most of the time

Previous setup: 4-6 Opus agents on high/max effort consumed tokens at similar rate

Issue duration: Started around June 15-17, continuing through June 18 (also June 13 showing no data but I have been using Claude Code almost nonstop for almost 2 weeks).

Despite drastically reducing usage (fewer agents, lower-tier model, reduced effort), consumption rate remains unchanged.

Potential tracking issue: Console Usage page shows zero tokens for June 13, 17, and 18, despite continuous Claude Code usage. June 14-15 show some API usage spikes, but don't reflect actual Claude Code activity.

Impact: Hit 100% of weekly limit despite minimal configuration. Even with a single idle session open, tokens continued depleting.

MCP servers active: 2 servers, Context Mode (which saves me token usage compared to not using it) and Claude Mem enabled (native)

What Should Happen?

Investigate please.

Error Messages/Logs

Steps to Reproduce

How can I reproduce this issue? ahah.

Claude Model

Not sure / Multiple models

Is this a regression?

I don't know

Last Working Version

_No response_

Claude Code Version

claude 2.1.177

Platform

Other

Operating System

Windows

Terminal/Shell

Other

Additional Information

I'm not using an API. This is Claude Code with tokens. Is this the right place to report this issue? I'm using Google Antigravity also.

View original on GitHub ↗