How to check your Codex usage: dashboard, /status and the apps
The quickest place to check your Codex usage is the usage dashboard at chatgpt.com/codex/settings/usage, which shows your current limits and when they reset. In the Codex CLI, /status shows what’s left of your 5-hour and weekly windows, /usage shows your token activity and redeems earned resets, and /statusline keeps the limits in the footer. In the ChatGPT desktop app, /status shows rate limits and /usage opens usage and billing. Business and Enterprise admins have workspace analytics.
Limits moved this week. On October 29, 2026, OpenAI doubled usage for all Pro tiers and Business Premium, two days after a reset for everyone. Your dashboard shows what applies to you; for what the limits are by plan, see Codex usage limits explained.
- The usage dashboard
- In the Codex CLI
- In the ChatGPT desktop app
- VS Code, iPhone and the web
- For Business and Enterprise admins
- Reading the numbers
- Resets: earned, paid and global
- Third-party trackers
- Questions
The usage dashboard
Open chatgpt.com/codex/settings/usage while signed in to ChatGPT. It shows your current limits and reset times, and OpenAI calls it the authority: the message counts on its pricing page are only estimates. The pricing page is the brochure; the dashboard is the meter. It’s also where you see your credit balance and buy credits, the same as Settings → Usage in ChatGPT, which lists any earned resets you have with their expiry dates. OpenAI suggests checking it every week or two to get a feel for your pace.
In the Codex CLI
/statusshows the session: model, approval policy, writable folders and token usage, and, when you’re signed in with ChatGPT, what’s left of your windows. Users’ reports show it as lines like5h limit: 89% left (resets 18:44)andWeekly limit: 82% left (resets 16:21 on 19 May), plus your credit balance if you have one. On a Pro plan there’s no 5-hour line. Since version 0.166.0 (October 29, 2026) it also names the ChatGPT workspace you’re using./usageopens a menu for your account’s token activity, or goes straight to it with/usage daily,/usage weeklyor/usage cumulative. The same menu redeems an earned reset./statuslinelets you pick footer items, including the rate limits, and saves them inconfig.toml, so you see what’s left without asking.
It’s natural to type /usage first. In the CLI that shows tokens, not limits; the limits are in /status. If /usage asks you to sign in, the session is using an API key, which is billed per token and has no windows to check.
In the ChatGPT desktop app
The Codex app became part of the ChatGPT desktop app on July 9, 2026. In a Codex chat there:
/statusshows the chat ID, how full its context is, and your rate limits./usageopens usage and billing settings, when you’re signed in with ChatGPT.- Usage widgets show what’s left, and adapt to accounts that have only a weekly limit.
On Enterprise and Edu, OpenAI’s help center also describes personal analytics in Settings → Usage & billing, with your usage over 7 or 30 days grouped by surface, model, reasoning effort or speed, which is the easiest way to see what’s using it up.
VS Code, iPhone and the web
- The IDE extension’s documentation doesn’t describe a usage view, and users have reported the limit display coming and going between versions. Use the dashboard or the CLI.
- On iPhone, since October 22, 2026 the ChatGPT app shows why you’re limited and when it resets right in the chat.
- On the web, eligible personal and Business accounts see their remaining usage in the ChatGPT sidebar. ChatGPT Work and Codex share the same limits and credits, so it’s one number for both.
For Business and Enterprise admins
- Codex analytics at admin.openai.com/analytics/codex covers Codex usage and Code Review across the workspace.
- Workspace analytics at chatgpt.com/admin/usage gives the wider ChatGPT picture.
- Usage insights in Admin Console → Analytics → Insights splits messages, credits and active users by use case.
- The Analytics API reports daily ChatGPT, Codex and Work usage, with an admin key that has the
enterprise.analytics.usage.readscope. - Billing on credit-based plans shows the credit balance and alerts, and lets admins set limits for the workspace, groups or individuals.
What you see depends on your role and the features turned on for the workspace.
Reading the numbers
- Two windows. Plus and Standard Business have a 5-hour window and a weekly one; whichever runs out first stops you, and the 5-hour one can run out before five hours pass. Pro plans currently have no 5-hour limit, only the weekly one.
- The weekly reset is yours. OpenAI publishes no common reset day; your dashboard and
/statusshow the time for your account. - Everything shares one allowance. Local messages and cloud chats draw from the same limits, and cloud tasks may use more of them. Reviews outside GitHub count as general usage.
- Some settings multiply usage. Fast mode counts 2.5 times and Ultrafast 8 times against included usage, image generation 3 to 5 times, and every MCP server adds context to each message.
- “Max” is a reasoning effort, not a model or a plan: it gives the model more time to think about one task, and uses more of your allowance.
- Credits come last. They’re spent only after your included limits run out.
For what uses the limits up and how they differ by plan, see what uses it up faster.
Resets: earned, paid and global
- Earned resets come from referrals and promotions. Redeem one with
/usagein the CLI or from Settings → Usage. OpenAI’s help center says a full reset refreshes both windows and moves your weekly reset date, and counts as used only once it refreshes a window. - Paid weekly resets are described in OpenAI’s help center for some personal plans: one restores both windows at once and restarts your week from your next request. We found no published price.
- Global resets are separate: OpenAI resets everyone’s limits from time to time and announces it on X. The latest was announced on October 27, 2026 by OpenAI’s Thibault Sottiaux and landed at 11:29 PM Pacific time (06:29 UTC on October 28). Sites that track these list a dozen in October alone; they track the global resets, not your own window.
OpenAI support can’t reset your limits, though it will look into usage that seems miscounted.
Third-party trackers
Several open-source tools read Codex’s local session logs and show usage without opening the dashboard. They aren’t OpenAI’s, and their dollar figures are API-price estimates, not your bill:
- ccusage: daily token use and estimated cost from local logs, for Codex as well as Claude Code.
- CodexBar: a macOS menu bar app with limits and reset times for Codex and other agents.
- codex-ratelimit: the 5-hour and weekly percentages and reset times from
~/.codex/sessions.
Seeing it without checking
We built Meowdoro, a focus timer around the MacBook notch that shows your Codex and Claude Code sessions, so you don’t have to keep opening the dashboard. It shows Codex’s 5-hour and weekly limits as one small pie for the fuller window. The pie turns orange from 80%, or earlier when, at your pace, the window would run out before it resets, and says when. Machines signed in to the same ChatGPT account share one group and its limits (how it shows agents). It also tells you when a Codex session finishes, so you don’t spend the window waiting for it. Try it and tell us what’s missing.
Questions
Where is the Codex usage dashboard?
At chatgpt.com/codex/settings/usage, signed in with the ChatGPT account you use for Codex. Settings → Usage in ChatGPT shows the same limits.
How do I check my Codex usage in the CLI?
Type /status for what’s left of your windows, or /usage daily, /usage weekly or /usage cumulative for token activity. /statusline can keep the limits on screen.
Why don’t I see a 5-hour limit?
Pro plans currently have only a weekly limit. On Plus the 5-hour window was lifted in July 2026 and came back in late August, so older screenshots may differ from yours.
Is there a Codex family plan?
No. Codex comes with ChatGPT Free, Go, Plus, Pro, Business, Enterprise and Edu, and OpenAI lists no family plan.
What is Codex Max?
A reasoning effort setting, not a model or a plan. It lets the model think longer about a task and uses more of your limits. The plan with the most usage is Pro at $500 a month.
When does Codex usage reset?
The 5-hour window resets five hours into it and the weekly one at a time set for your account; both times are on the dashboard and in /status. OpenAI also runs occasional global resets.
How does this compare with Claude Code?
Claude Code has the same two windows and shows them in /usage. See Claude Code vs Codex usage limits.
Related guides
Sources: OpenAI’s Codex pricing documentation, the slash command references for the CLI and the ChatGPT desktop app, models, workspace analytics, the changelog and what’s new; OpenAI help center articles on Codex with a ChatGPT plan, resets and credits; posts on X by Thibault Sottiaux of OpenAI on October 27 and 29, 2026; user reports in the Codex issue tracker for the /status output. Read on October 30, 2026.