If you pay for ChatGPT and also use Codex, OpenAI’s coding agent, your usage limits are about to change. OpenAI has updated its help centre to say that from December 2026, ChatGPT and Codex will draw from one shared usage allowance on certain paid plans. The details are still thin, so here is what is confirmed, what is not and what you should do now.
Quick answer: On Plus, Pro and Business plans, ChatGPT and Codex will share a single allowance measured in tokens. Heavier models and agent tasks use more of it, while quick Instant chats do not count. Your combined usage and reset times will appear in Settings. OpenAI has not yet given the exact date or the size of each plan’s allowance.
Note: based on OpenAI’s help centre updates from late October and early November 2026, as reported by tech outlets.
What OpenAI has confirmed
- One shared allowance across ChatGPT and Codex on Plus, Pro and Business plans.
- Token-based counting: heavier models, long conversations and agent tasks use more.
- Instant chats are not counted against the allowance.
- A combined usage view, with reset times, will appear in Settings.
What is still unknown
- The exact date in December.
- How large the allowance is for each plan.
- Whether the current five-hour and weekly limits stay.
- How Free, Go, Enterprise and Edu plans are affected, since they were not named.

Before and after
| Before December | From December | |
|---|---|---|
| Limits | Separate limits for ChatGPT and Codex | One shared allowance |
| How usage is counted | Mostly by messages or tasks | By tokens |
| Instant chats | Counted in some limits | Not counted |
| Where to check | Different places | One usage view in Settings |
Who should pay attention
- Light users who mostly use quick chats will probably notice little difference.
- Heavy users of advanced models, long research tasks or Codex coding agents could hit the shared limit sooner, because everything comes from one pool.
- Teams on Business plans should check how shared usage affects their daily work.
What to do now
- Note how often you hit limits today, in ChatGPT and in Codex.
- Use Instant mode for quick questions, since it will not use your allowance.
- Check Settings for the new usage view once the change goes live.
- Avoid buying an annual plan based on guesses about the new limits.
For practical ways to stretch your allowance, read how to make ChatGPT usage limits last longer. OpenAI’s help centre will have the official details as they are published.
The bottom line
From December, ChatGPT and Codex share one token-based allowance on Plus, Pro and Business plans. Most casual users will not notice much, but heavy users should track their usage and lean on Instant mode for simple questions.
Frequently asked questions
When will ChatGPT and Codex share usage limits?
OpenAI says December 2026, but has not given an exact date.
Which plans are affected?
Plus, Pro and Business plans were named. Free, Go, Enterprise and Edu were not mentioned.
Do Instant chats count towards the shared limit?
No. OpenAI says Instant chats do not draw from the shared allowance.
Where can I see my usage?
OpenAI says a combined usage view with reset times will appear in Settings.


