openai / openai/codex

Allow users to consume the weekly Codex quota continuously instead of enforcing the 5-hour cap

Open
#43,135 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

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

Description

What variant of Codex are you using?

app

What feature would you like to see?

Feature request

Please allow users to consume their available weekly Codex quota continuously, or provide an optional "weekly-pool mode" where the 5-hour rolling limit does not interrupt an active engineering task.

Problem

The current 5-hour limit makes long autonomous coding sessions difficult even when a large amount of weekly quota is still available.

My typical Codex workload is not a short coding question. It often involves:

  • auditing a large repository
  • modifying multiple services
  • running tests
  • deploying to GCP / Vercel
  • checking browser behavior
  • inspecting logs
  • fixing failures
  • redeploying
  • repeating acceptance tests
  • producing evidence and commits

A single serious task can require several hours of continuous agent work.

With high-compute models, the 5-hour bucket can be exhausted very quickly during these deploy → test → fix → retest loops. The task then stops even though substantial weekly quota may still remain.

This is especially frustrating because the model may be in the middle of diagnosing or fixing the exact blocker when the short-window quota expires.

Preferred behavior

I would prefer to be able to spend my weekly allowance faster if I choose.

For example:

  • Weekly allowance remaining: 50%
  • 5-hour allowance remaining: 0%

Instead of forcing the task to stop until the 5-hour reset, Codex could optionally continue drawing from the remaining weekly allowance.

Possible implementations:

  1. Optional "Use weekly quota continuously" mode
  2. Allow an already-running task to continue from the weekly pool
  3. Let users choose between the current 5-hour pacing and an unrestricted weekly-pool mode
  4. Offer a warning before consuming the weekly quota rapidly, rather than enforcing a hard stop

Why this matters

For long-running autonomous engineering work, continuity is often more valuable than spreading usage evenly across the week.

If I choose to use most of my weekly quota in one day to complete a production milestone, I would rather have that option than be forced to stop several times during the same task.

The previous behavior where weekly usage could effectively be consumed more continuously was much more useful for this kind of workflow.

Environment

  • Codex desktop/app
  • Long-running repository and infrastructure tasks
  • High-compute models
  • Large multi-service codebase
  • Frequent terminal, browser, deployment, and test tool usage

This is not primarily a request for more total weekly compute.

It is a request for more control over when the existing weekly compute can be consumed.

Additional information

No response

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

The issue identifies the Codex app and describes weekly and 5-hour quota behavior, but it names no source files, tests, or implementation entry point. Start by locating the quota-enforcement component and its tests, then determine whether an optional weekly-pool mode can be specified and verified without changing total weekly allowance.

Written by the indexing model from the issue text.

Assessment

Domain
backend
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.