ChatGPT and Codex will share one usage limit from December

Plus, Pro and Business plans move to a single token based allowance across chat and coding, with Instant chat left out.

OpenAI says that starting in December 2026, Plus, Pro and Business subscribers will have one shared usage allowance for ChatGPT and Codex. Work in both products will count against the same token based limit, Instant chat will not count, and the remaining balance will be visible in Settings, under Usage.

Key facts

OpenAI will merge ChatGPT and Codex usage into one shared allowance for Plus, Pro and Business plans, starting in December 2026.
Usage is measured in tokens, which vary by model and task, so advanced models and complex jobs use more.
Instant chat will not draw from the shared allowance, according to OpenAI as quoted by TokenPost.
Combined remaining usage and reset times appear in Settings, under Usage.
OpenAI has not given an exact start day or published allowance amounts for each plan.

What happened

The change was reported on October 31 and shared on the OpenAI Developer Community forum on November 1, in a post that links to a new OpenAI help center article on ChatGPT and Codex usage limits. The post says that from December, allowances on the Plus, Pro and Business plans will be shared between ChatGPT and Codex, as part of the ongoing work to bring Work into Chat. It also says OpenAI expects most users will not need to wait for their usage to reset more often than they do now.

TokenPost adds the practical details. Activity in both products will count against one token based limit. Tokens depend on the model and the task, so heavier models and longer jobs use more of the balance. Instant chat is the exception: OpenAI says it will not draw from your usage. When you reach the limit, you can wait for the reset, move to a higher plan, or buy credits for eligible features.

Codex is OpenAI’s coding agent, available inside ChatGPT, in code editors and in the terminal. TokenPost notes that subscribers had already seen a notice about the planned change at checkout, but without a start date or plan allowances, and those numbers are still not public.

Why it matters

Until now, many subscribers treated ChatGPT and Codex as two separate budgets: a long coding session did not touch the chat allowance, and the other way round. From December, they become one budget. If you use both, a big refactoring job in Codex in the morning can leave less room for research or writing in ChatGPT in the afternoon.

For people who mostly chat, the exclusion of Instant chat matters. Quick everyday questions that run on Instant will not eat into the shared pool, so the allowance is mainly spent on heavier reasoning, agent work and coding.

The change also shows where OpenAI is heading. Chat, agents and coding are turning into one product with one meter, and the meter is tokens, not messages. That makes planning more important: knowing which tasks are heavy, when your limit resets and where to check it will save you from hitting a wall in the middle of a deadline. Because exact numbers are not out yet, it is worth checking Settings, under Usage, during the first weeks of December to see how fast your own work uses the balance.

How to try it

The simplest way to get ready is to plan your week so that heavy AI work does not pile up on one day. Our free Plan my week prompt turns your commitments and goals into a realistic plan with buffers and one clear priority per day. Fill in the fields like this:

Fixed commitments: “Team call Monday 10:00, client review Wednesday 15:00, school pickup every day at 16:30”
Goals for the week: “Finish the invoice export feature with Codex, write the monthly report, research two competitors”
Free hours per day: “3”
When you work best: “Morning”

When you get the plan, check three things. First, that the heavy jobs, like the coding feature and the research, sit on different days instead of the same morning. Second, that each day has a buffer, so a reset wait or a long agent run does not push everything back. Third, that light tasks, such as quick questions and short rewrites, are placed where you can handle them with Instant chat. If the plan puts two big jobs together, ask the chatbot to move one of them and explain the new order.

Try it

Plan my week

Frequently asked questions

When do ChatGPT and Codex start sharing one usage limit?

OpenAI says the shared allowance starts in December 2026 for Plus, Pro and Business subscribers. It has not announced a specific day, and the per plan allowance amounts have not been published yet.

Does Instant chat count toward the shared ChatGPT and Codex limit?

No. According to OpenAI, as quoted by TokenPost, Instant chat will not draw from your usage. The allowance is spent by heavier models and tasks in ChatGPT and by work in Codex, measured in tokens.

What can I do when I hit the shared usage limit?

You can wait for the allowance to reset, upgrade to a higher plan, or buy credits for eligible features. You can see the combined remaining usage and the reset times in ChatGPT under Settings, then Usage.

Sources