openai / openai/codex

Urgent: apparent zero remaining usage shortly after reset; two reset cards reported within ~1 hour (macOS)

Open
#44,244 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

app bug rate-limits
Dominant language
Rust
Stars
125k
Forks
19.4k
PR merge metrics
PR metrics pending

Description

Severity / summary

Urgent investigation requested: the user reports redeeming two Codex usage-reset cards within approximately one hour, with the UI appearing to show zero remaining usage shortly after a reset. This disrupted ongoing work.

This is a user-reported symptom, not a confirmed billing or redemption failure. A subsequent read-only allowance check did not show the account exhausted, so the discrepancy needs investigation rather than being presented as proven loss of credits.

Environment
  • Desktop app on macOS
  • Desktop version: 26.903.61454, build 8378
  • Bundled CLI: codex-cli 0.153.4
  • Reported during the night of September 9–10, 2026
User-reported sequence
  1. Redeem a usage-reset card.
  2. Continue working.
  3. Observe an apparent zero-remaining-usage state unexpectedly soon.
  4. The user reports using two reset cards within approximately one hour.

Exact redemption timestamps, redemption receipts, and the affected UI counter were not captured. Work was ongoing; this is not a claim that no usage occurred.

Independently observed current state

A later read-only allowance response returned:

  • Codex primary allowance: usedPercent = 0, windowDurationMins = 10080
  • Additional credit balance: balance = "0", hasCredits = false
  • Available usage-reset credits: availableCount = 1
  • spendControlReached = false
  • rateLimitReachedType = null

The primary allowance being 0% used must not be confused with 0% remaining, an additional-credit balance of 0, or the token counter of a newly created goal. The particular counter that the user saw is not yet established.

The local application logs inspected did not expose both redemption receipts. No additional reset was redeemed during this investigation. Current allowance data alone cannot establish whether the two earlier redemptions were applied correctly or whether usage was charged correctly between them.

Expected behavior
  • Each successful reset clearly updates the relevant allowance.
  • Usage allowance, extra-credit balance, reset-card count, and goal token counters have unambiguous labels.
  • If usage falls rapidly after a reset, there is enough dated usage/redemption history to explain it.
  • A stale, absent, or differently scoped quota bucket should not appear as an unexplained exhausted allowance.
Requested investigation

Please investigate whether this could involve a stale allowance display, bucket selection, delayed usage attribution, reset application, or another issue. We are not asserting any one of these as the root cause.

Please provide a private support route if account-specific redemption and metering records are needed; do not request account identifiers or payment evidence in this public issue.

Potentially related, not confirmed same cause: #42765.

Privacy

The user explicitly approved a sanitized public report. No account identifiers, email addresses, private chat contents, project/source files, server details, passwords/tokens, reset-card IDs, or payment receipts are included. Raw logs are not attached publicly.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

No source file, test, or code entry point is identified. Start by distinguishing the UI counter from the allowance, extra-credit, reset-card, and goal-token fields, then compare the read-only allowance response with dated redemption and usage records through the private support route. Done means the affected counter and any stale bucket, attribution delay, or reset-application discrepancy are identified and explainable.

Written by the indexing model from the issue text.

Assessment

Tech stack
macos, rust
Domain
backend, desktop
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.