ChatGPT and Codex Will Share Usage Limits: Does Regular Chat Count?
Fact-checked: November 2, 2026
OpenAI plans to combine the usage allowances for ChatGPT and Codex from December for Plus, Pro, and Business plans.
The question for many users is simple: Will chatting with ChatGPT use up the allowance available for Codex? Could frequent chats leave less room for coding?
According to OpenAI's explanation:
- Instant chat will not draw from the shared allowance.
- ChatGPT and Codex will share an allowance.
- Usage will be measured in tokens, not simply by message count.
- Consumption can vary by model and task.
- The explanation does not list specific allowance amounts or the full calculation rules.
OpenAI says the change starts in December but does not specify a year or exact date. Given the date of this fact check, we refer to it here as expected in December 2026.
Does regular ChatGPT chat use the shared allowance?
It depends on the model and feature you use.
Instant chat: no shared allowance deducted
OpenAI says Instant chat will not draw from the combined allowance. If you mainly use Instant chat to answer questions, translate, or edit text, those chats will not use the shared allowance. The explanation does not define the boundaries for additional tools used in a chat, so check the rules in effect for other features.
Not drawing from the shared allowance does not mean unlimited use. Instant may still have other limits under your plan.
Advanced reasoning: watch for the details
If you frequently use advanced reasoning settings such as Medium or High, keep an eye on the change. OpenAI says ChatGPT and Codex will share a token-based allowance and that consumption depends on the model and task. It does not specify how Medium and High are each counted.
For example, you might use ChatGPT to analyze a complex problem in the morning and Codex to change code in the afternoon. If the earlier chat counts toward the shared allowance, the two tasks may use the same pool. A demanding earlier task could leave less for coding.
For now, you cannot calculate exactly how much one High chat uses or determine from this explanation alone how every reasoning setting is counted.
Will the shared allowance run out faster?
It could for some users, but that is not certain for everyone.
Under the December plan, some ChatGPT usage and Codex usage will be counted together. Heavy use of both eligible chat and Codex could therefore affect the allowance available for either task.
Imagine researching and analyzing information in ChatGPT every day while also using Codex to write and revise code. If that research counts toward the shared allowance and uses a substantial amount, less may remain for Codex during the current allowance period.
There is already some sharing: ChatGPT Work and Codex currently share a usage allowance. It would be misleading to say all ChatGPT features previously had limits entirely separate from Codex. Whether users reach the limit sooner depends on the new allowance and how different tasks are counted.
OpenAI expects most users will not need to wait for a reset more often than they do now. That is an expectation, not a guarantee for every user. It is too early to conclude that Plus or Pro has been given a smaller allowance.
Who should pay closer attention: Plus or Pro users?
The practical impact depends on how you work, rather than your plan alone.
| Usage pattern | What to watch |
|---|---|
| Mostly Instant chat | Instant chat does not draw from the shared allowance; check the rules for other features |
| Frequent Medium or High reasoning | Check whether each setting counts and how usage is calculated |
| Mostly coding with Codex | See how much coding the new allowance supports |
| Heavy use of chat and Codex | Watch whether eligible chats and coding tasks affect each other's available usage |
| Business plan | Wait for details about plan and team limits |
People who rely on both advanced chat and Codex have the most reason to pay attention. Some use reasoning models to analyze technical plans or review code; others research a problem in ChatGPT before asking Codex to act. These tasks often work together.
In an OpenAI community discussion, some users worried that a shared pool for research and coding could disrupt their workflows. These are individual concerns, not evidence that everyone will be affected. The practical impact will be clearer after the rules launch.
What does measuring the shared allowance in tokens mean?
The number of messages is not the only thing that matters: the model and the task matter too.
Tokens are units of information an AI model processes or generates. A simple question and an analysis of extensive material may each take one message, but they generally require different amounts of resources. Likewise, changing a few lines of code with Codex and handling a complex project may consume very different amounts.
OpenAI says the model and task affect consumption, but its explanation does not provide a complete formula. For now, you cannot determine:
- How much shared allowance one High chat will use.
- How much a Codex coding task will use.
- How many complex tasks Plus and Pro will each support.
Do not use OpenAI API token prices to calculate a ChatGPT subscription allowance. The billing and usage rules differ and cannot be converted directly.
What happens when the allowance runs out?
OpenAI says users can wait for a reset, upgrade their plan, or buy credits to continue using eligible features.
It also says the combined remaining usage and reset times will appear under ChatGPT → Settings → Usage. Once the change is live, that is a useful place to check whether your allowance meets your needs. The current usage page should not be assumed to be the final shared-allowance interface.
Should you upgrade from Plus to Pro now?
There is no need to upgrade solely because the allowances are being combined.
OpenAI's explanation does not yet answer several key questions in full:
- How much shared allowance will Plus and Pro each have?
- How will usage be calculated for different models?
- How often will the new allowance reset?
- How much eligible advanced chat will remain after heavy Codex use?
- How will the new allowance compare with today's limits in practice?
Without those answers, it is hard to judge whether an upgrade is worthwhile. If you mainly use Instant chat, you can continue as usual. If you rely heavily on Codex or both advanced chat and Codex, note whether you currently hit limits and decide after the new rules launch and you can see your actual usage.
What remains unclear about the December change?
As of November 2, 2026, OpenAI has described its plan but has not included every detail in the explanation.
| Question | What the explanation says |
|---|---|
| When will sharing begin? | Starting in December; no year or exact date stated. Based on the fact-check date, December 2026 is expected |
| Which plans are involved? | Plus, Pro, and Business |
| Does Instant chat count? | No; the explanation does not define the boundary for additional tools |
| How much will Plus and Pro receive? | Specific amounts are not listed in the explanation |
| How will different models be counted? | Complete rules are not listed in the explanation |
| How often will it reset? | The explanation does not give a specific reset period for the new shared allowance |
| Does this mean a lower allowance? | It is not yet possible to tell |
These details matter when deciding whether to change plans. Avoid making that decision based on unverified allowance figures.
Frequently asked questions
When will ChatGPT and Codex share an allowance?
OpenAI says the change for Plus, Pro, and Business starts in December. It does not state a year or exact date; based on the date of this fact check, December 2026 is expected.
Will Instant chat use my Codex allowance?
No. OpenAI says Instant chat does not draw from the new shared allowance. Its explanation does not define the precise treatment of additional tools.
Will ChatGPT High affect Codex?
If advanced reasoning chats count toward the shared allowance, they could affect what remains for Codex. The explanation does not give High's specific calculation rules, so the cost of one chat cannot be determined.
Does sharing make Plus less worthwhile?
It is too early to tell. Wait for the allowance amounts and consumption rules before comparing actual usage before and after the change.
Will I still need to buy Codex separately?
For the plans covered by this change, OpenAI plans to include ChatGPT and Codex in a shared allowance. Whether you need extra credits depends on your usage and the rules in effect; the merger alone does not answer that question.
Takeaway
The effect will depend on how you use ChatGPT and Codex. If you mainly use Instant chat, OpenAI has explicitly said it will not draw from the shared allowance.
If you rely on advanced reasoning and Codex for coding, watch the applicable rules and your usage after the change. It is too early to say the allowance will necessarily run out faster or that Plus or Pro has been reduced.
For now, keep your current plan and revisit the decision when the details are available and you can assess your own usage.
Sources
- OpenAI: Understanding ChatGPT and Codex usage limits — the planned shared allowance, Instant chat, and where to check usage.
- OpenAI: ChatGPT Work and Codex pricing — existing shared usage and subscription rules.
- OpenAI community discussion — concerns raised by some users.