The OpenAI logo. Image: OpenAI (logo, public domain via Wikimedia Commons)
Your ChatGPT conversations are about to come out of the same pot as your Codex coding sessions. OpenAI says in a new help centre article that “starting in December, Plus, Pro, and Business plans’ usage allowance will be shared across ChatGPT and Codex”, as it brings “the agentic capabilities of Work into the main ChatGPT experience”.
The page appeared on Saturday, October 31, and quickly climbed Hacker News. Subscribers there say OpenAI has also emailed them about it, under the headings “Chat and Work are coming together in one experience” and “A shared allowance for ChatGPT and Codex”. OpenAI hasn’t given a firm date in December.
What changes in December
Today, ChatGPT Work and Codex already draw on one allowance, as our ChatGPT Work explainer sets out, while ordinary chats run on their own separate limits. From December, chat joins the shared pool. The help page says:
- Tokens, not messages. Usage will be “measured in tokens, the units of information a model processes and generates”. How much each request costs depends on the model you pick and the task, and “more advanced models and complex tasks generally use more of your allowance”.
- Instant chat is free of the meter. “Instant chat will not draw from your usage,” the page says, so quick everyday questions on the Instant model won’t eat into your coding time.
- One usage screen. You will see your combined remaining usage and reset times in Settings, then Usage.
- At the limit, you can wait for the reset, upgrade your plan or buy credits to keep using eligible features.
OpenAI says it is “preparing carefully for this transition to help prevent disruption to your workflows” and that it expects “most users won’t need to wait for their usage to reset more often than they do now”.
Who it hits
The people most exposed are those who lean on both products at once: developers who plan and review in ChatGPT with a heavy reasoning model, then build in Codex. Until now, a long research session in chat cost nothing against Codex. From December, every non-Instant request does, and a big Codex run could leave less room for chat.
That split was the selling point for some subscribers. A researcher posting on Hacker News as yurimo wrote:
one big reason I as a researcher have been a customer of OpenAI is that on their Pro plan Pro-tier model in chat and codex are decoupled in terms of billing
yurimo, Hacker News
Another user, furyofantares, said running out of Claude Code usage and then losing “much more mundane usage, queries or web searches” as well had been a reason to prefer ChatGPT and Codex. Anthropic already counts Claude chats and Claude Code against the same limits, and its top plan, now called Ultra, dropped its five-hour limit on October 30.
The change lands weeks after OpenAI cut Pro 200 from 20 times to 10 times the Plus allowance in Work and Codex, a cut that took effect on October 30. Codex users have also grown used to surprise global resets; OpenAI hasn’t said how those will work once chat shares the pool.
What OpenAI hasn’t said
The help page is short, and several things are missing:
- Numbers. It gives no token allowance for any plan, so there is no way yet to compare before and after.
- Message caps. It doesn’t say whether per-model caps in chat, such as Pro 200’s 100 GPT-6 Pro messages a week, go away or stay alongside the new pool.
- Other plans. It names only Plus, Pro and Business. Free, Enterprise and Edu aren’t mentioned.
- Instant. It doesn’t define which model counts as Instant once Work is folded into the main ChatGPT app.
OpenAI hasn’t published a blog post or statement beyond the help page. The page doesn’t single out any country, so the change looks set to apply wherever those plans are sold.
Why it matters
ChatGPT’s near-unlimited chat on paid plans has been one of OpenAI’s quiet advantages over Claude, and that ends for anything heavier than Instant. If the per-plan allowances turn out generous, most people won’t notice; if not, the developers who pay the most are the ones likeliest to look elsewhere.
Sources: OpenAI Help Center: Understanding ChatGPT and Codex usage limits; OpenAI Help Center: credits; Hacker News discussion.
Latest OpenAI news
