Codex is charging excessive credits for use, not consistent with 1 week ago

What version of Codex is running?

codex-cli 0.44.0

What subscription do you have?

Plus

Which model were you using?

gpt-5-codex

What platform is your computer?

Linux 6.16.10-200.fc42.x86_64 x86_64 unknown

What issue are you seeing?

Codex is consuming usage limits much faster than just a few days ago.
I ran something like 5 or so tasks before it told me I had reached the 5 hour limit... this is a radical change from the last time I used the service just a week ago. If you look at my screen shot, notice my usage on oct 22... this did not trigger any limits... but here nov 5, showing a sliver of use by comparison, triggered the limit. I even purchased the extended credits... and it seems to be charging me something like 1-2$ per task... (i managed to use up nearly 200 credits in just over 1 hour).

This seemed a bit extreme, so I checked the forums... and it seems it is not just me...
https://community.openai.com/t/codex-using-up-massive-credits-850-credits-5-hour-limit-used-on-only-8-queries/1364774/16
https://community.openai.com/t/pro-plan-hit-5-hour-limit-twice-in-2h-and-1-5h-and-nearly-exhausted-weekly-cap-in-1-day-after-today-s-update/1364782/20

This affects both codex CLI and web

What steps can reproduce the bug?

Codex CLI

1) run codex
2) dispatch a task
3) open codex on web
4) click settings
5) click usage
6a) see that usage percentage for 5 hour period and 7 day period are using far more proportion than expected.
OR 6b) see that far more credits are consumed than expected.

Web

1) Login to open ai
2) go to codex
3) dispatch several tasks
4) click settings
5) click usage
6a) see that usage percentage for 5 hour period and 7 day period are using far more proportion than expected.
OR 6b) see that far more credits are consumed than expected.

What is the expected behavior?

<img width="1100" height="316" alt="Image" src="https://github.com/user-attachments/assets/feabe981-f124-4d39-b743-72a630720c61" />

Additional information

_No response_