Skip to content

Weekly quota draining insanely fast #21050

Open

Description

@bcdonadio

What version of Codex CLI is running?

0.128.0

What subscription do you have?

Pro USD$200/mo

Which model were you using?

gpt-5.5

What platform is your computer?

Linux 6.19.12-100.fc42.x86_64 x86_64

What terminal emulator and version are you using (if applicable)?

Warp v0.2026.04.29.08.57.stable_01

What issue are you seeing?

My weekly usage rate in the Pro 20x plan is draining insanely faster than it was before the gpt-5.5 launch. The 5h quota seems to be unaffected, however. Besides, the status line is not updating the weekly quota.

I'm doing the same usual work as before. I was never ever able to get below 80% of my weekly quota even with very heavy usage. I'm now already at 19% remaining only and this wasn't even a heavy use week.

What steps can reproduce the bug?

Uploaded thread: 019df3bb-3c5f-7861-b6c2-736d4dac1bc1

What is the expected behavior?

Usage quota compatible to the charged cost and intended professional heavy use.

Additional information

No response

Activity

  1. added
    CLIIssues related to the Codex CLI
    rate-limitsIssues related to rate limits, quotas, and token usage reporting
    TUIIssues related to the terminal user interface: text input, menus and dialogs, and terminal display
    on May 4, 2026
  2. removed
    TUIIssues related to the terminal user interface: text input, menus and dialogs, and terminal display
    CLIIssues related to the Codex CLI
    on May 4, 2026
  3. bcdonadio commented on May 4, 2026

    @bcdonadio
    Author
    Image

    In the breakdown above (back to February 4th) you can see that my utilization barely registered until the day of gpt-5.5 release, which then skyrocketed with my same normal use and now even avoiding the fast mode for a lot of sessions.

    The most important fact is the launch day: it is exactly when it skyrocketed.

  4. bcdonadio commented on May 4, 2026

    @bcdonadio
    Author

    One more information: I always used gpt-5.4 xhigh with 1M context window and in fast mode. That meant a 4x burn rate before gpt-5.5 with the extra high thinking.

    Now I'm using just a few tasks on fast, lowered to just high, and my quotas are in levels I've never seen before.

  5. jodoe69135 commented on Jun 21, 2026

    @jodoe69135

    I’m seeing the same issue on a Pro plan x20.

    My usage pattern has not materially changed: same project, same model, same reasoning effort, and if anything I have been using Codex less than usual recently.

    I have been using Codex for weeks/months without hitting the weekly limit, and usually with plenty of weekly quota remaining. Recently, however, the weekly quota is being exhausted much faster.

    From my local Codex logs, after deduplicating token_count events because forked/resumed sessions can reimport old events:

    • Previous observed weekly window, reset June 19: usage went from ~2% to ~69% over ~5 observed days, without hitting the weekly limit.
    • Next weekly window, reset June 25: usage went from 0% on June 18 09:00 to 100% on June 21 16:54.
    • That means the weekly limit was exhausted in roughly 3 days and 8 hours, with almost 4 days still remaining before the scheduled reset.

    This does not match my historical usage at all.

  6. sergey-telegin commented on Aug 16, 2026

    @sergey-telegin

    At first, Codex consumed my entire weekly usage limit in a single day, even though my workload and usage pattern were essentially the same as before.

    After that, I purchased an additional $20 of Codex usage. That entire $20 balance was consumed in roughly 10 minutes of normal work.

    This is drastically different from my previous experience with Codex. I did not significantly change my workflow, project, or usage intensity. The rate at which both the weekly allowance and the additional paid usage were consumed seems abnormal.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingrate-limitsIssues related to rate limits, quotas, and token usage reporting

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions