Quick answers
Can I calculate my remaining Codex allowance from token usage?
Not directly. Local token records and official account allowance may use different sources and time windows. Check the official usage dashboard for your remaining allowance, and local records for past token usage. A missing record does not mean no usage, and an old reading may no longer be accurate. Worked example.
For your current account limits, start with OpenAI.
OpenAI's Codex plan guidance explains where to view current usage limits in ChatGPT settings or the usage dashboard. In an active Codex CLI session, /status can also show the current session's usage-limit information. The exact entry point and fields can vary by account and product version, so use the live official view for the current answer.
Peilo is useful when you also want to review locally available activity and work history. It is not a substitute for the account's authoritative limit status.
Four numbers that are easy to confuse.
| Measure | What it answers | What it does not prove |
|---|---|---|
| Account limit or allowance | What usage is currently available to this account, and when does the limit reset? | A local token total is not a reliable conversion to remaining account allowance. |
| Token activity | How much input and output activity was recorded for the available sessions? | A local record may not cover every device, account, or session. |
| Cost estimate | What would recorded token activity cost under a selected pricing model? | An estimate is not an invoice or a statement of subscription charges. |
| Work history | Which projects or topics appear in the available conversations? | Conversation counts do not equal work hours, productivity, or completed tasks. |
ChatGPT plan allowances and API billing are different systems. API rate limits or API prices should not be used to infer a ChatGPT plan's Codex allowance. Check the account and product context before comparing.
OpenAI API documentation: API rate limits → OpenAI API pricing →What the Peilo computer app can help you review.
Depending on your app version and connected services, the dashboard shows session status, recent token usage, usage history, and available account limits. Work Process groups local conversation records into summaries and collaboration patterns.
These panels have different sources and time windows. A value may be unavailable, stale, or absent because the source could not be read or no matching local history exists. When a reading is missing or out of date, do not treat it as zero usage.
AI-assisted interpretation is optional. Treat a generated explanation as a prompt for review, then verify it against the original conversation. It is not proof of task completion, focus, or work quality.
Read the computer-app guide → Review one AI collaboration →If two numbers do not match, check the context.
- Account and source: are both views about the same account, device, and data source?
- Time window: is one number a current account limit while the other covers a local history period?
- Coverage and freshness: did local history load, and when was the account source last updated?
Missing history, a failed read, an offline account service, a stale value, and zero usage are different states. Keep the source label and error details when asking for help. Do not clear local data just to make totals look alike.
Local records are not a full account ledger.
The computer app reads the local sources you authorize. It may not include activity from other devices or accounts. Optional AI interpretation can send selected text excerpts to the provider you configure; review the product's privacy explanation before enabling that feature.
Never share an API key, account credential, private conversation, or unredacted dashboard screenshot in a public support request.
Read Peilo’s privacy notice →Worked example: check the source before the number.
Illustrative example: local records show 1.2 million tokens in the past hour; the official account page shows 84% remaining in its current window. These answer different questions. You cannot convert the token total into remaining allowance. A ten-minute interruption is a gap, not zero usage; a stale 84% reading is not the current allowance.
Check the account and workspace, time window, source coverage, and update time. Use the official Codex usage dashboard for account limits; /status is available in an active CLI session. Plan, model, and allowance rules can change.
Official source checked October 6, 2026. Using Codex with your ChatGPT plan
About this page
This guide explains the difference between local usage records in Peilo PC and your Codex account limits. Plans, models, and limits can change; check your official account page and the sources linked in this guide.
Maintained by AnyoneKit · Updated October 6, 2026