ChatGPT and Codex will share one usage limit from December

OpenAI is changing how paid ChatGPT plans count usage. According to a new OpenAI Help Center article, from December the usage allowance on Plus, Pro and Business plans will be shared across ChatGPT and Codex, the company's coding agent. OpenAI says the change comes as it brings the agentic capabilities of its Work mode into the main ChatGPT experience.

Tokens instead of message counts

Usage for both products will be measured in tokens, the units of text a model reads and writes. How quickly you use up your allowance will depend on two things:

• Which model you pick. More advanced models use more of the allowance.
• What you ask it to do. Long, complex tasks, like a multi-step coding job in Codex, cost more than a quick question.

One important exception: Instant chat will not draw from your usage. Everyday quick chats in ChatGPT should therefore stay unaffected, while heavier work in ChatGPT and Codex competes for the same pool.

What OpenAI promises

OpenAI says it is preparing the transition carefully and expects that most users won't need to wait for their usage to reset more often than they do now. You'll be able to see your combined remaining usage and reset times under Settings → Usage.

If you do hit the limit, the options are to wait for the reset, upgrade your plan or buy credits to keep using eligible features.

What we don't know yet

The help article doesn't give an exact start date in December, the size of each plan's allowance or how many tokens typical tasks will use. If you use Codex heavily on a Plus plan, it's worth watching the usage page once the change arrives, since coding sessions and long ChatGPT tasks will now eat into the same budget.

Context

The change follows a busy month for OpenAI, including reports that it cut its funding target by $20 billion. If you use ChatGPT for work, also check our guide to ChatGPT privacy settings. Developers who would rather avoid usage caps altogether can look at running open-weight AI models locally.