One allowance across chat and code
Starting in December, OpenAI says the usage allowance for Plus, Pro and Business subscribers will be shared across ChatGPT and Codex. Usage will be measured in tokens, with consumption varying according to the selected model and the complexity of the task.
The announcement establishes a common allowance for eligible activity across both products. It does not provide a complete comparison with today's plan-specific limits or say that every existing usage restriction will disappear. For subscribers who rely on advanced ChatGPT models and Codex, the practical question is how their combined workload will fit within the new allowance.
OpenAI frames the change as part of bringing Work's agentic capabilities into the main ChatGPT experience. It does not announce that Codex is being discontinued or merged into ChatGPT as a product. The confirmed change is to usage accounting: eligible activity across ChatGPT and Codex will draw from a shared allowance.
Instant is exempt. The allowance size is unknown.
OpenAI says Instant chat will not consume the shared allowance. That protects ordinary Instant conversations from this particular usage meter, but the documentation does not explain every boundary between Instant chat and more advanced or delegated tasks.
The allowance itself remains unspecified. OpenAI gives no token amount for Plus, Pro or Business, no exact December start date, and no reset schedule for the new shared system. It says consumption will depend on the model and task, meaning subscribers cannot yet calculate how many typical Codex or advanced ChatGPT sessions their plan will support.
What OpenAI has confirmed
OpenAI's Help Center now explicitly describes the December transition. It identifies the affected plans, the token-based measurement, the Instant chat exemption and where subscribers will see their remaining usage.
It also says it expects most users will not have to wait for a reset more frequently than they do today. That is OpenAI's expectation, not an observed result of the new system.
The documentation leaves the commercially important details open: allowance sizes, exact implementation date, reset cadence and the price of additional credits. Until those are published, subscribers cannot independently evaluate whether the shared allowance will provide more or less capacity for their own workloads.
Sources
- [1] OpenAI Help Center, “Understanding ChatGPT and Codex usage limits” (article 20001625), read first-hand: the December sharing sentence, token measurement and its dependence on the model chosen and the task, the Instant exemption, combined remaining usage and reset times in Settings → Usage, the options at the limit (wait for the reset, upgrade the plan, or buy credits for eligible features), and the expectation that most users will not need to wait for a reset more often than they do now; the page publishes no allowance size, no December day, no reset cadence and no credit priceRead source
- [2] Automater, “ChatGPT and Codex Move to One Shared Token Allowance in December”, October 31, 2026: an independent capture of the same help-page text, with the page's October 31 update stampRead source