OpenAI to Merge ChatGPT and Codex Into One Shared Usage Limit
OpenAI is about to change how its paying subscribers use two of its most popular products. Starting in December, ChatGPT and Codex will draw from one shared usage allowance for Plus, Pro and Business plans, replacing separate limits for chatting and coding with a single token-based pool, according to OpenAI’s Help Center.
The change was set out in OpenAI’s Help Center article “Understanding ChatGPT and Codex usage limits,” updated on October 31, 2026, and picked up in reporting over the weekend and on Monday, November 2. It is a practical shift for millions of American users who move between asking ChatGPT questions and asking Codex, OpenAI’s coding agent, to write, review or fix software. From December, heavy work in one product will leave less room in the other.
OpenAI has not published an exact start date in December, nor has it published the size of each plan’s allowance. That missing detail is itself part of the story: subscribers know the shape of the change, but not yet how large their shared pool will be. What OpenAI has said, in its own words, is that “starting in December, Plus, Pro, and Business plans’ usage allowance will be shared across ChatGPT and Codex.”
What changes in December
Today, a paying subscriber effectively has more than one meter to watch. Chat use and coding-agent use are tracked separately, and users who hit one limit may still have room in the other. From December, that separation ends for the three main paid tiers. Conversations in ChatGPT and coding work in Codex will both count against the same allowance.
Usage will be measured in tokens, the small units of text that AI models read and write. OpenAI says more advanced models and more complex tasks generally use more of the allowance. A long, detailed coding session or a demanding analysis task will therefore burn through the pool faster than a short, simple exchange.
There is one clear exception. OpenAI says “Instant chat will not draw from your usage.” Instant chat is the fast, default chat experience, and it will stay outside the shared meter. Everything else that counts — chats with advanced models and Codex agent work — will share the same balance.
Subscribers will be able to see the combined position in one place. OpenAI says combined remaining usage and reset times will appear in Settings, under Usage. When the allowance runs out, users will have three options, according to the Help Center: wait for the reset, upgrade their plan, or buy credits to continue using eligible features.
Why OpenAI is pooling the limits
OpenAI frames the change as part of bringing agentic capabilities into the main ChatGPT experience. Its Help Center says the shared allowance arrives “as we bring the agentic capabilities of Work into the main ChatGPT experience.” In other words, the line between chatting with an AI and handing it a job to carry out is blurring, and OpenAI is now metering that work as one thing rather than two.
That logic is already visible elsewhere in OpenAI’s plans. On Plus and Pro plans, Codex, ChatGPT Work, Excel and PowerPoint already share one agentic allowance when those features are available, according to OpenAI’s documentation. December extends that pooling principle to chat with advanced models, so a subscriber’s everyday questions and their coding-agent sessions sit in the same budget.
For OpenAI, a single token pool is also simpler to explain and to operate than a patchwork of separate caps for each product. For users, however, simplicity comes with a trade-off: a busy afternoon of Codex work can no longer be treated as separate from an evening of ChatGPT research. The two now compete for the same resource.
Who will feel it most
The subscribers most likely to notice the change are those who use both products heavily. Developers and small teams who run Codex on a Plus, Pro or Business plan are the clearest example. A long debugging session, a large code review or an automated refactoring job can consume a large number of tokens, and under the new system that consumption will reduce what is left for ChatGPT chats until the next reset.
Business users face a similar planning problem. Teams that use ChatGPT for drafting, analysis and research, and Codex for building and maintaining software, will need to think of their subscription as one shared budget rather than two independent ones. That may mean staggering heavy coding jobs, choosing a lighter model for routine questions, or tracking usage more closely through the Settings dashboard.
Casual users are less likely to hit the ceiling. OpenAI says it expects most users will not need to wait for their usage to reset more often than they do now. That is a forecast, not a guarantee, and it will depend on how large the unpublished allowances turn out to be and on how intensively each subscriber uses the advanced models.
What to check before December
Subscribers do not need to take any action today, but a few simple checks will make December smoother. First, open Settings and then Usage in ChatGPT to see how the dashboard presents remaining usage and reset times. Second, if you use both ChatGPT and Codex, start noting which kind of work uses the most — long Codex runs, advanced-model chats, or both — so you can spot pressure on the shared pool early.
Third, remember the Instant chat exception. For quick, routine questions, Instant chat will not draw from the shared allowance, which makes it a useful fallback when the main allowance is running low. Fourth, plan for the three options at the limit: waiting for a reset costs nothing but time, upgrading raises the ceiling, and buying credits offers a way to keep working without changing plans. OpenAI has not yet published credit prices for December overflow use, so subscribers should treat that option as confirmed in principle but unpriced for now.
It is also worth watching for a fuller announcement. So far, the change has surfaced through the Help Center rather than a blog post, and reporting on Monday noted that no separate OpenAI blog post or social post announcing it had been found. If OpenAI publishes allowance sizes, an exact date, or credit pricing before December, those details will matter more than the headline itself.
Part of a wider AI pricing reset
The shared allowance arrives at a moment when the economics of AI are under close scrutiny. Big Tech companies are spending hundreds of billions of dollars on data centers, chips and power, and investors are asking how quickly that spending will pay off. NationWire reported on November 2 that the Magnificent Seven’s combined debt had jumped 45 percent to $578 billion by mid-2026, driven by AI infrastructure spending.
OpenAI itself has been adjusting its commercial terms in recent months. It cut its fundraising target from $30 billion to $20 billion after some investors trimmed commitments, according to reports covered by NationWire on November 1, amid growing investor caution about AI spending. A shared, token-based allowance fits that broader pattern: AI companies are moving toward metering actual compute use more precisely, rather than offering open-ended access at a flat price.
For American subscribers, the practical question is not whether AI pricing will keep changing — it almost certainly will — but whether the new system is transparent enough to plan around. A single pool is easier to understand than separate caps, but only if users can see how much they have left, when it resets, and what extra credits will cost. On those three points, OpenAI still has details to publish before December.
Frequently Asked Questions
What is changing for ChatGPT and Codex in December?
For Plus, Pro and Business plans, ChatGPT and Codex will share one usage allowance, measured in tokens. Work in either product will count against the same pool, instead of separate limits for chat and coding.
Which plans are affected?
Plus, Pro and Business plans. OpenAI’s Help Center names those three tiers for the December change.
Does Instant chat count against the shared allowance?
No. OpenAI says Instant chat will not draw from your usage, making it an exception to the shared meter.
How is usage measured?
In tokens, the units of information a model processes and generates. More advanced models and more complex tasks generally use more of the allowance.
What happens when I reach the limit?
According to OpenAI, you can wait for the usage to reset, upgrade your plan, or buy credits to continue using eligible features. Combined remaining usage and reset times will be shown in Settings, under Usage.
Has OpenAI published the exact date or allowance sizes?
Not yet. OpenAI says only “starting in December” and has not published a specific day or per-plan allowance amounts.
Sources
The Neuron: Everything That Happened in AI Today, Monday, November 2
Precise Impact AI: AI Daily Brief, November 2, 2026
Automater: ChatGPT and Codex Move to One Shared Token Allowance in December
Crypto Briefing: OpenAI to merge ChatGPT and Codex token allowances in December 2026
Related coverage
Magnificent Seven Debt Hits $578 Billion on AI Spending
OpenAI Cuts Fundraising Target to $20 Billion as AI Investors Turn Cautious
AI Giants Agree on One Opt-Out: How Publishers Can Block AI Training With a Single Setting
