Codex usage. In your menu bar.

Check your usage now.

Already in the Codex CLI? Run /status during your session to check remaining limits. You can also use the official usage dashboard linked in OpenAI’s plan guide. This reads status; it does not start a coding task.

What you can see.

INMC reads available plan limits from the signed-in Codex account on this Mac. The menu bar uses weekly remaining capacity when reported; the detail view keeps a reported five-hour window separate. Each reset time belongs to its own window. Limits vary by plan, and some plans have no five-hour cap. Eligible tools can share allowances, so do not assume each surface has an independent quota.

The percentage is remaining plan capacity. It is not an API balance, a token count or a prediction of how many tasks you can finish. The year calendar describes recorded token activity, with missing days left empty.

Get started.

1. Sign in to Codex on this Mac with the account you want to monitor.
2. Open INMC and choose Codex. Refresh once to request a current reading.
3. Read the reported limit windows and their reset times. Check the last-reading time before planning a long task; a five-hour window does not apply to every plan.

One useful workflow.

1. Start with one concrete outcome, such as fixing a failing checkout test.
2. Provide the reproduction, relevant files and the expected behavior. Ask Codex to identify the smallest useful change.
3. Review the diff and the agreed check before expanding the task. A smaller task improves clarity; it does not guarantee a fixed amount of savings.

If something does not match.

• If the reading is missing, confirm that Codex is signed in on this Mac. Unknown never means 100% available.
• If the account changed, refresh after signing in to the intended account. Compare the same limit window and reading time.
• API-key use and ChatGPT plan use are different billing surfaces. Check the official plan guide before comparing them.