To check Codex usage, open the official usage dashboard. In an active Codex CLI session, enter /status to see your remaining limits. Read each allowance together with its reset time: a token counter answers a different question.
Here’s how to find the right number, keep it visible while you work, and optionally give it a place on your desk.
Check your Codex usage dashboard
- Open the Codex usage dashboard and sign in with the account you use for Codex.
- Find the current allowance windows and their reset times.
- Read whether each percentage means used or remaining. Keep the window label with the number.
Use the limits shown for your account. Plans can have different windows, and published message counts are estimates. The model and work involved affect consumption. OpenAI’s usage guidance points to the dashboard for current limits and resets.
Check usage in the Codex CLI
Inside an active Codex terminal session, type:
/statusThis opens the session summary. It includes information such as the active model and token usage; OpenAI also documents it as the way to check remaining limits in the CLI. Read the labels before comparing a number with the browser dashboard.
For a Codex usage monitor that stays in your terminal, enter:
/statuslineUse the picker to choose and order the footer items you want. Rate limits, context information and token counters are separate choices. Put rate limits in the footer if your main question is how much allowance remains. See the official status-line instructions.
Enter these commands in Codex, not at an ordinary shell prompt. If your client doesn’t offer a command, the browser dashboard is the direct fallback.
What the different usage numbers mean
| Reading | What it tells you | Use it for |
|---|---|---|
| Allowance used or remaining | Your position within one quota window. | Deciding when to start another substantial task. |
| Reset time | When that particular window next resets. | Planning around the limit you’re approaching. |
| Context information | The conversation’s context capacity. | Understanding the current session. |
| Token activity | Recorded processing activity over the stated scope. | Reviewing activity, rather than calculating allowance left. |
| API spend | Money reported for API usage. | Tracking an API project or organization’s costs. |
For example, 68% remaining means 32% used in that same window. It doesn’t mean 68% of a fixed token balance or 68% of your subscription payment is left. Don’t add a weekly percentage to a shorter-window percentage.
For builders, OpenAI exposes allowance data through account/rateLimits/read. It supplies separate bucket identifiers, used percentages, window durations and reset timestamps. Token activity has a separate account/usage/read method. Missing fields should stay unknown. The App Server reference documents the distinction.
Keep Codex allowance visible on TickrMeter
A terminal footer works while you’re looking at the terminal. TickrMeter’s AI Usage display gives the allowance a small, persistent home beside your keyboard. One fewer tab to go hunting for.
The personal Codex connection uses a local helper on the computer where Codex is installed and signed in. TickrMeter provides the display app. You don’t need to build a custom collector or connect an MCP client for this ready-made route.
- Open Apps in TickrMeter, choose AI Usage, and use the personal Codex / ChatGPT connection.
- Follow the portal’s connection instructions. Run its one-time helper command on your signed-in Codex computer.
- Check the connection, then use the personal usage options available to your account.
- Preview the setup and compare its allowance label and value with the same window in Codex before sending it to your chosen TickrMeter.
The helper needs an awake, connected computer to keep sending readings. A sleeping laptop cannot supply a fresh snapshot. Check the source and connection status when a value appears stuck; a retained reading is not proof that nothing has changed.
Available connections and display options are shown in your setup. The TickrMeter integration guide explains the app and local-bridge routes. This personal setup is for allowance and reset information; choose the separate organization API connection when you want API spend.
If the readings don’t match
- Check the account. The browser and the Codex installation used by the helper should be signed into the intended account.
- Check the window. Compare the same named allowance and reset, not whichever percentage looks most familiar.
- Check the direction. “Used” increases as allowance is consumed; “remaining” decreases.
- Check freshness. Reopen the source reading and inspect the connection before treating an older display value as current.
Start with the dashboard, choose the reading that helps you plan, and keep that one in view. Explore TickrMeter if you’d like it beside your screen.
