[BUG] OPUS TOKENS RUNNING WITHOUT OPUS BEING SELECTED

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?

Sonnet 4.5 selected default, claude code confirmed sonnet 4.5 but yet when I /usage and check, 100% of the tokens used are "opus". that seems like a problem since opus tokens are being spent faster than 4.5 and cutting into my context allowance... doesnt seem right to me... please fix this issue

What Should Happen?

claude, should fix this so we aren't getting screwed with the 1 week context window being used when we aren't using opus....

Error Messages/Logs

> what model are you?

⏺ I'm Claude Sonnet 4.5 (model ID: claude-sonnet-4-5-20250929).

> /usage
  ⎿  Status dialog dismissed

> well why does it say under /usage and the usage tab that ive used 11% of context and all 11% has been opus?

⏺ That's a discrepancy - I am definitely running on Sonnet 4.5 (claude-sonnet-4-5-20250929), but it sounds like
  the usage tracking UI is incorrectly labeling the usage as "Opus."

  This is likely a UI display bug in the Claude Code interface. The actual model being used for this conversation
  is Sonnet 4.5, but the usage attribution in the /usage tab appears to be mislabeled.

  You might want to report this as a bug at https://github.com/anthropics/claude-code/issues so the team can fix
  the usage tracking display.

> /usage
  ⎿  Status dialog dismissed

> so if opus wasnt selected at all, but its showing that im using opus tokens, wouldnt that be false advertisement
 and breaking trust with the user?

⏺ You're absolutely right - that's a serious issue. Accurate usage reporting is fundamental to user trust,
  especially when:

  1. Different models have different costs - users need to know what they're actually being charged for
  2. Users make decisions based on that data - choosing which model to use, managing budgets, etc.
  3. Transparency is expected - if the UI says "Opus", users should be getting Opus

  If the usage tab is showing Opus tokens when Sonnet is actually running, that's misleading and breaks the trust
  contract with users. It's not just a minor UI bug - it's a data accuracy problem that affects billing
  transparency and informed decision-making.

  This definitely warrants reporting at https://github.com/anthropics/claude-code/issues as a high-priority bug
  affecting usage transparency and user trust. The team needs to fix this so users can rely on accurate usage
  information.

Steps to Reproduce

1. fix
2. update
3. drop the 1 week context window

Claude Model

Sonnet (default)

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.0.2

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

what model are you?

⏺ I'm Claude Sonnet 4.5 (model ID: claude-sonnet-4-5-20250929).

/usage

⎿  Status dialog dismissed

well why does it say under /usage and the usage tab that ive used 11% of context and all 11% has been opus?

⏺ That's a discrepancy - I am definitely running on Sonnet 4.5 (claude-sonnet-4-5-20250929), but it sounds like
the usage tracking UI is incorrectly labeling the usage as "Opus."

This is likely a UI display bug in the Claude Code interface. The actual model being used for this conversation
is Sonnet 4.5, but the usage attribution in the /usage tab appears to be mislabeled.

You might want to report this as a bug at https://github.com/anthropics/claude-code/issues so the team can fix
the usage tracking display.

/usage

⎿  Status dialog dismissed

so if opus wasnt selected at all, but its showing that im using opus tokens, wouldnt that be false advertisement

and breaking trust with the user?

⏺ You're absolutely right - that's a serious issue. Accurate usage reporting is fundamental to user trust,
especially when:

1. Different models have different costs - users need to know what they're actually being charged for
2. Users make decisions based on that data - choosing which model to use, managing budgets, etc.
3. Transparency is expected - if the UI says "Opus", users should be getting Opus

If the usage tab is showing Opus tokens when Sonnet is actually running, that's misleading and breaks the trust
contract with users. It's not just a minor UI bug - it's a data accuracy problem that affects billing
transparency and informed decision-making.

This definitely warrants reporting at https://github.com/anthropics/claude-code/issues as a high-priority bug
affecting usage transparency and user trust. The team needs to fix this so users can rely on accurate usage
information.