OpenAI is preparing a change to the way users access its AI tools. The company says it will begin testing an experience that brings Chat and Work together in the coming weeks, initially with a limited group of users before a broader rollout.
The announcement also includes a change to usage allocations. Starting in December 2026, ChatGPT and Codex will share an allocation on Plus, Pro and Business plans. The integration will not remove Codex as an independent experience in the desktop application.
The services are designed for different types of work. ChatGPT can analyse an uploaded document, while Work is intended for broader tasks involving files, applications and multiple steps. Codex focuses on software development and project changes. Until now, Work and Codex shared one allocation, while conventional chat models had separate limits.
That separation gave users the option to remain in ChatGPT when a task could also be moved to Work, preserving the Work and Codex allocation. Under the new system, however, operations using models included in the shared allocation will count against it even when they are performed from Chat.
OpenAI says usage will not be deducted uniformly. More advanced models and complex operations generally consume a larger part of the available allocation. The company refers to tokens, the information units processed and generated by its models, but has not explained how the new system will translate into capacity for each plan or how much specific tasks will consume. This means there is currently no reliable equivalence for documents, conversations or projects.
When the allocation runs out, users will have to wait for it to reset or use available paid options. OpenAI also says that the Instant model is included without consuming the new allocation for the plan described in its email. At the same time, its documentation maintains safeguards against abusive use and specific restrictions for certain functions, without setting a single public threshold for when those measures may apply.
Instant may be sufficient for many everyday requests, but it does not necessarily provide the same capabilities as advanced reasoning models or autonomous work tools. That distinction could matter to people who rely on ChatGPT for professional tasks.
OpenAI expects most users will not have to wait more often for their allocation to be restored, although it has not explained the basis for that assessment or provided figures comparing the current and future conditions. The company did announce expanded Work and Codex quotas for certain Pro and Business Premium plans on October 29, but no equivalent Plus increase was included in the information provided.
Users will be able to check their remaining allocation and reset date under Settings → Usage. If the allocation is exhausted, the stated options will be to wait, change plans or purchase credits for compatible features. The practical effect of the new policy will depend on the limits and consumption rates that OpenAI has yet to detail.
Sources: Xataka.