Show remaining usage limit in Codex app chat
Resolved 💬 6 comments Opened Apr 27, 2026 by QuiltyVal Closed May 18, 2026
💡 Likely answer: A maintainer (etraut-openai, contributor)
responded on this thread — see the highlighted reply below.
Feature request
Please bring back a visible usage / remaining limit indicator directly under the Codex app chat, like it was before.
Right now the limit is too opaque: users can run into usage limits while actively working without enough warning, and the remaining quota/reset time is not visible in the place where the work is happening. That makes Codex harder to trust for longer development sessions.
Expected behavior:
- Show current usage or remaining messages/percentage in the Codex app/chat UI.
- Show reset time for relevant windows.
- Keep it visible near the chat/status area without requiring a separate settings page or command.
Related but different from #15281, which focuses on CLI /status//statusline. This request is specifically for the Codex app chat UI.
6 Comments
Which variant of Codex are you using? CLI, app, IDE extension, web?
Although I am not the OP, I use both the CLI on macOS and the desktop app on Windows, and I dread having to click into this[^1] menu. Would love to have it display both the 5h and 7d usage near the context progress, with a hover popover that shows the details like reset time.
[^1]: <img width="291" height="367" alt="Image" src="https://github.com/user-attachments/assets/66cb8acc-f377-43f4-b3ff-db3f11a4cfc8" />
Using the Mac desktop app, I don't see anywhere this is displayed...am I just missing it???
0.131.0-alpha.9 Feedback:
It's absolutely bogus you omitted this feature, I wouldn't have updated and paused all my work just so I can roll back. It's absolutely and irrefutably asinine to force this feature and call the reasoning "psychological ux".
These are results for "why did codex windows take away usage limit view in chat":
OpenAI removed the visible usage limit indicator from the chat and UI to reduce user anxiety, shifting focus away from micro-managing tokens. The change also accommodates the shift toward rolling, token-based usage and variable context-window scaling.
OpenAI Developer Community
The primary reasons behind the removal include:
Psychological UX: Developers have noted that keeping a running tally causes "budget anxiety". OpenAI removed these counters so casual users aren't pressured to use up every last drop of their allowance, and power users don't stress about tracking marginal tokens.
Shift to Token-Based Accounting: Since the introduction of token-based pricing for Business and Enterprise tiers, counting literal "messages" is no longer accurate. Usage now drains based on the size of the repository, background context syncs, and the specific reasoning demands of models like GPT-5.4.
Dynamic Compression: Because modern agents dynamically compact previous conversation threads to save space rather than just hitting a hard cutoff, fixed message counters became obsolete.
OpenAI Developer Community
How to Check Your Limits Now
Although the inline chat meter is gone, you can still view your remaining quota:
In the CLI: Run the /status command inside your active session.
Via Web Dashboard: Visit the OpenAI Platform Usage Dashboard for tokens and costs, or check the Codex Pricing page for tier metrics.
Custom Widgets: Advanced users can create custom tracking widgets to dock in their taskbar or sidebar.
My bad, I found where usage/limits are currently shown directly in the Codex app, so I am closing this issue.
That said, I still do not understand why UI elements need to keep moving around. The remaining limit is important during active Codex work, so it would still be useful to show usage and reset time directly in the Codex app chat/status area, where the user is already working, instead of moving it to a less obvious place.
I am experiencing the same issue with the WEB interface, which completele conceals the fact that work-labeled chats consume paid quota tokens, and have also created a separate post about it here:
Stop Hiding Weekly Usage: ChatGPT Work Now Drains the Codex Quota Without Warning
The lack of a clear, persistent usage indicator—combined with the poorly explained shared quota between Codex and ChatGPT Work—is causing unnecessary confusion for paying users.
I am linking the discussions so the reports can be cross-referenced and the broader impact of this issue is easier to see.