OpenAI Plans Shared ChatGPT and Codex Usage Limits for December

OpenAI’s planned usage change will make everyday ChatGPT conversations and Codex coding work compete for a common allowance. That could make its products feel more unified, but subscribers won’t know what the shift means for their practical capacity until OpenAI explains how large the pool will be and how it will count consumption.
The change is scheduled to begin in December for Plus, Pro and Business subscribers, according to a help article published over the weekend. OpenAI says more demanding models and complex tasks will use a larger share of the combined token allowance. Its faster Instant chat mode, intended for everyday questions and tasks, won’t draw from that pool.
The missing figure is the one subscribers need most: OpenAI hasn’t said how generous the shared allowance will be. The company says most people should see usage reset no more often than they do now, but without a comparison against current limits, that assurance is difficult to assess for people who use both products heavily.
One account, more shared capacity
The proposal extends a product consolidation already underway. OpenAI’s agentic tools have been moving closer to ChatGPT, giving users a way to delegate work such as research, data analysis and the creation of documents, presentations and spreadsheets alongside conversational tasks.
In July, OpenAI launched ChatGPT Work, built on its Codex coding technology. The product broadened Codex beyond software development, positioning agents as tools for knowledge workers handling business tasks. At the time, The New Stack reported that OpenAI had also combined ChatGPT and Codex in one desktop application, with separate interfaces for conversation, agentic tasks and coding.
Other products have added to the mix. ChatGPT for Excel debuted in March and later extended to Google Sheets. Codex, ChatGPT Work, ChatGPT for Excel and Workspace Agents already share an agentic usage allowance. Regular ChatGPT conversations currently have their own limits. In December, OpenAI plans to bring those conversations into the shared arrangement.
That change matters most to users who move between distinct kinds of work. A developer might use Codex intensively to build or review software, then turn to ChatGPT to research a technical question or plan the next phase. Under a common cap, spending in one product could leave less capacity for the other.
Subscribers want clearer accounting
OpenAI’s direction follows a broader push to let users move between conversation and delegated work in one interface. In September, Anthropic merged Claude Chat and Cowork, enabling users to shift from asking questions to assigning more involved tasks within the same conversation. For both companies, the product logic is straightforward: reduce the friction of switching tools as a task grows more complex.
Usage limits add a harder question. When those capabilities draw on one allowance, users need to understand what consumes it and how much remains. A discussion on OpenAI’s developer forum called on the company either to retain separate limits or guarantee that subscribers won’t lose usable capacity.
“ChatGPT and Codex serve different purposes, and their usage limits should remain separate.”
The forum discussion’s starter, posting as Smalikou, argued that the products serve different needs and asked OpenAI to explain how it would calculate usage. Another participant, PZA, focused on visibility: a useful dashboard, they argued, should break down consumption by product, model and task, while showing the remaining allowance and reset time. They wrote that without such transparency, they would see the change as a downgrade and reconsider renewing.
“Without that transparency, I will consider the change a downgrade to my paid plan and reconsider renewing.”
OpenAI has said users will be able to see their combined remaining allowance and reset times in ChatGPT settings. It hasn’t specified whether those settings will show the product, model and task breakdown subscribers are asking for. That detail will determine how easily users can spot which activities are consuming their capacity.
Recent limit changes shape the response
Subscriber concerns arrive soon after OpenAI reversed a proposed cut to the $200 Pro plan. On October 29, the company said it was cancelling the reduction, restoring higher Work and Codex allowances for subscribers already moved to lower limits, and doubling allowances for Pro 100, Pro 500 and Business Premium subscribers.
Tibo Sottiaux, OpenAI’s head of ChatGPT and Codex, said on X that the company had heard subscriber feedback and would keep Pro 200 usage limits intact. “We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same,” he wrote.
Some forum participants question whether higher limits will translate into more room once ChatGPT conversations also count against the allowance. One user, Zigg88, said they already regularly approached their weekly Codex limit. “Doubling a resource allocation while simultaneously expanding the number of things consuming it doesn’t inherently translate into more usable capacity,” they wrote.
Other users raised concerns about using ChatGPT for research and technical planning while relying on Codex for software development. Their proposed remedies include keeping separate allowances or guaranteeing equivalent access after the change.
The New Stack asked OpenAI for the exact December start date and whether the shared pool would match the capacity provided by today’s separate limits. The company declined to answer those questions, instead pointing to its recent Work and Codex increases and existing documentation.
OpenAI’s consolidation may make it simpler to move between chat and delegated work, but a single pool also makes capacity harder to judge without clear accounting. The practical test will be whether the published allowance, its consumption rules and the usage display give subscribers enough information to plan their work across both products.
The post OpenAI’s new shared usage limit pits ChatGPT against Codex appeared first on The New Stack.