How to check Codex token usage: dashboard, /status and /usage

To check Codex token usage, open the usage dashboard at chatgpt.com/codex/settings/usage, or type /status inside a Codex CLI session. OpenAI's Codex pricing page names both as the places to see where you stand, and that is as much as the official docs promise.

One thing to know before you look: OpenAI frames Codex usage as limits and credits, not as a running token count. The docs explain that tokens are the units of text the model reads and writes, and that your prompt, files, chat history, tool results and the response all use them. What they point you to is your remaining allowance.

Where to check Codex usage

All three are described in OpenAI's docs as of October 27, 2026. The command reference does not list subcommands or version requirements for them, so if a command is missing in your install, update Codex first and check the reference again.

How to check your usage step by step

1. Sign in with your ChatGPT account and open chatgpt.com/codex/settings/usage. This is the page the docs call the usage dashboard.
2. Read your current limits and when they reset. Plus and Standard Business plans have a five-hour window, and weekly limits may also apply. The docs say Pro plans currently have no five-hour limit. OpenAI says the estimates on its pricing page are not fixed, so trust the dashboard over any number you remember.
3. In a running Codex CLI session, type /status to see the remaining limits and the context usage for that chat.
4. Type /usage if you want the usage and billing settings opened for you. The command reference says this works when you sign in with ChatGPT.
5. Repeat the dashboard check every week or two. OpenAI recommends that rhythm to track your pace.

What counts as tokens in Codex

Everything the model sees. OpenAI lists the prompt, the files, the chat history, the tool results and the reply. That is why a long chat costs more per message than a fresh one: each turn carries the history with it. For the difference between the size of one chat's memory and your plan's allowance, see our explainer on the Codex context window and the usage limit.

What happens when you hit the limit

The docs say that once you reach your included limits, available credits let you continue working. Plus and Pro users can buy extra credits, and Business, Edu and Enterprise workspaces with flexible pricing can buy workspace credits. Credits are the unit used to pay for eligible usage on credit-based plans. Purchase prices depend on your plan or agreement.

How to make your usage last longer

OpenAI's pricing page gives five tips:

• Keep prompts lean. Be precise, and remove unnecessary context.
• Narrow the source material. Give Codex only the relevant files.
• Scope the output. Say who it is for, the format and the length, and separate required work from optional improvements.
• Trim your AGENTS.md files. On a larger project, nest them in the repository to control how much context is injected.
• Disable MCP servers you are not using. Each one adds context to your messages and uses more of your limit.

If usage runs higher than expected, the docs also suggest asking whether a smaller model or a tighter task scope would still give a useful result.

What the official pages do not show

The dashboard and /status answer "how much is left". Neither is described in the docs as a history of tokens by day or by model. If you want that view, local dashboards exist, and we compared them with OpenAI's page in Codex usage dashboard: OpenAI's page vs a local view. Anything a third-party tool shows is an estimate. For billing questions, use your OpenAI account.