OpenAI is moving ChatGPT toward a single place for asking questions and getting larger tasks done. The company is testing a combined Chat and Work experience that would let people move from an ordinary conversation to research, file-based projects, presentations, websites, and browser tasks without switching modes. At the same time, the ChatGPT shared usage limit is changing: Plus, Pro, and Business subscribers will draw from one token-based allowance across ChatGPT and Codex starting in December.
The two changes are connected, but they are not the same rollout. The interface test is expected to begin with a small group in the coming weeks, according to OpenAI's email to subscribers, as reported by TechCrunch and BleepingComputer. The shared usage policy is documented in OpenAI's Help Center notice.
Chat and Work are moving into one ChatGPT experience
The planned test combines Chat and Work so users can start with a question, explore an idea, and then ask ChatGPT to carry out a more involved task in the same experience. OpenAI says people in the test will be able to bring in files and connected apps, ask ChatGPT to build a website or create a presentation, and delegate browser-based work while reviewing and guiding the process.
That changes the role of Work. Instead of feeling like a separate mode users must remember to open, its research and task-execution capabilities become part of the main ChatGPT flow. The distinction between a quick answer and a longer assignment becomes less about choosing a tab first and more about what the user asks ChatGPT to do.
OpenAI has not said the combined interface is available to everyone. The initial test is limited, and the company has not announced a general release date. Codex also remains a separate experience in the desktop app; the reported change does not mean the coding product is being folded into the Chat tab.
The ChatGPT shared usage limit will include Codex
The more immediate change for subscribers arrives in December. OpenAI says Plus, Pro, and Business plans will share one usage allowance across ChatGPT and Codex. Usage will be measured in tokens, the units models process and generate, rather than as a simple count of prompts.
The amount consumed will depend on the model and the task. OpenAI says advanced models and complex tasks generally use more of the allowance. That makes a long research assignment, a multi-step browser task, or a substantial coding session potentially more demanding than a short exchange.
OpenAI says Instant chat will not draw from the shared ChatGPT and Codex allowance. The shared pool applies to the usage covered by the new policy, not every message sent in ChatGPT.
What subscribers can expect when they reach the limit
OpenAI says users will be able to see their combined remaining usage and reset times under Settings → Usage. If they reach the allowance limit, the listed options are to wait for the reset, upgrade their plan, or buy credits for eligible features.
The company expects most users will not need to wait for a reset more often than they do now. That is OpenAI's expectation, not a guarantee for every subscriber. The Help Center notice does not publish exact token allocations for each plan, a per-task price list, or the precise December start date.
The change also makes plan details more relevant to people who use ChatGPT for everyday questions and Codex for software development. Saganote's ChatGPT pricing guide explains the subscription tiers, while its ChatGPT Work vs Claude Cowork comparison looks at the broader category of AI systems designed to complete work rather than only answer prompts. Readers comparing assistants can also refer to the ChatGPT vs Gemini vs Claude guide.
For teams managing access to these tools, Saganote has also covered OpenAI's admin plugin for ChatGPT Work and Codex. The shared allowance does not replace workspace permissions or administration, but it adds another resource that paid subscribers and administrators may need to monitor.
OpenAI is joining conversation and execution, not merging every product
The direction is clear: ChatGPT is being shaped into a workspace where a conversation can become a finished deliverable. The interface test reduces the friction between asking and delegating, while the shared allowance puts more advanced ChatGPT tasks and Codex work under one usage budget.
For users, the practical change will depend on how often they rely on Work and Codex, which models they choose, and how demanding their tasks are. The interface test is still limited, and the shared usage policy is scheduled for December. Until OpenAI publishes plan-level token amounts and a firm start date, subscribers can confirm the scope of the change but cannot calculate exactly how much work their allowance will cover.
