OpenAI is folding ChatGPT and Codex usage into a single allowance. Starting in December, everything you do in ChatGPT and Codex on a Plus, Pro or Business plan will draw from one pool measured in tokens, according to a new OpenAI help center article. The only thing left outside it is Instant chat, the fast no-reasoning mode.
The change came out in emails to subscribers over the weekend, alongside news that OpenAI will start testing a merged Chat and Work tab with a small group of users. Work and Codex already share a limit today. What’s new is that the rest of ChatGPT joins them, and OpenAI hasn’t said how big the combined pool will be.
- Price: no change announced. Plus is still $20 a month, Pro is $100, $200 or $500, and Business is $25 per user monthly or $20 billed annually.
- When and who: “starting in December” on Plus, Pro and Business, with no exact date. OpenAI’s article doesn’t mention Free, Go, Enterprise or Edu.
- What counts: ChatGPT and Codex use, measured in tokens, with bigger models and harder tasks using more. Instant chat doesn’t count.
- The catch: chats that used to sit outside your Codex limit will now compete with it, and OpenAI hasn’t published the size of the new allowance. It only says it expects most people won’t hit resets more often than now.
What OpenAI says changes in December
The help article is short, and the key lines are these: “Starting in December, Plus, Pro, and Business plans’ usage allowance will be shared across ChatGPT and Codex as we bring the agentic capabilities of Work into the main ChatGPT experience.” Usage will be counted in tokens, the chunks of text a model reads and writes, and how fast you burn through the allowance “will depend on the model you choose and the task you ask it to do.”
OpenAI says you’ll be able to see combined remaining usage and reset times in Settings, under Usage. When you run out, the options are the ones Codex users already know: wait for the reset, upgrade your plan or buy credits.
The reassurance is a prediction, not a promise. “We expect most users won’t need to wait for their usage to reset more often than they do now,” the article says. TechCrunch, which saw the email, reports that OpenAI told subscribers some people might have to wait for resets more often, depending on usage and model.
How ChatGPT and Codex usage works today
Right now, ChatGPT has two kinds of limits. Chat has its own model limits, and Work and Codex share a separate agent allowance. OpenAI’s Codex pricing page says it plainly: “ChatGPT Work and Codex share usage.” Some add-ins already draw from that pool too. OpenAI’s release notes say ChatGPT for Excel uses the same agent limits as Codex on Plus and Pro, and that ChatGPT for Word draws from the shared allowance.
For Plus and Standard Business seats, OpenAI publishes rough estimates of how many local Codex messages fit in each five-hour window:
| Model | Local messages per 5 hours (Plus and Standard Business) |
|---|---|
| GPT-6 Astra | 5 to 45 |
| GPT-6.1 Sol | 15 to 160 |
| GPT-6.1 Terra | 35 to 285 |
| GPT-6 Sol | 15 to 150 |
| GPT-6 Luna | 350 to 3,000 |
OpenAI stresses that these are estimates, not fixed caps, that cloud tasks can use more, and that weekly limits may also apply. Pro plans currently have no five-hour limit. Speed settings matter as well: Fast mode uses included usage at 2.5 times the normal rate, and the Ultrafast tiers of GPT-6 Astra and GPT-6.1 Sol use it at 8 times. Image generation burns through limits about three to five times faster on average. We covered the Ultrafast pricing in our GPT-6.1 Sol Ultrafast breakdown.
What actually changes for you
Codex users lose a fallback. Today, if you max out Codex, you can still keep chatting in ChatGPT, because Chat lives under different limits. From December, a long research conversation and a coding session will come out of the same bucket.
The Instant exemption is the important detail. OpenAI’s article says “Instant chat will not draw from your usage,” and the email, as quoted by 9to5Mac, calls it “Unlimited Instant.” Because only Instant is named, our reading is that chats where you pick a reasoning level, such as Medium, High or the Pro options, will count. OpenAI hasn’t said exactly which chat modes draw from the pool, so treat that as likely rather than confirmed.
Pro subscribers have more headroom. On October 29, OpenAI doubled the Work and Codex allowance on Pro 100 and Pro 500, and it canceled a planned cut for Pro 200, bringing those subscribers up to the same higher limit. On Business, Premium seats cost $100 per user per month billed annually or $125 monthly, and OpenAI’s Business release notes describe them as having five times the Standard usage, no five-hour limit and weekly resets. The October 29 increase also covered Business Premium.
Plus is where people will feel it most. A $20 plan whose agent allowance already runs from 5 to 45 Astra messages per five hours now has to cover thinking-mode chat as well, unless OpenAI makes the pool bigger. It hasn’t said whether it will.
Chat and Work are merging, and ChatGPT Classic is going away
The usage change is tied to a design fix. Since July, the ChatGPT desktop app has had a switcher between ChatGPT and Codex, and inside ChatGPT a toggle between Chat and Work. OpenAI now plans to combine Chat and Work into one experience. According to the email quoted by TechCrunch, that will let you “move from exploring an idea to having ChatGPT carry it through, all in one place,” and “Codex will remain a separate experience in the ChatGPT desktop app.”
The combined tab starts as a test with a small group in the coming weeks, and OpenAI hasn’t given a date for everyone else. Separately, 9to5Mac reports that the older ChatGPT Classic desktop app now shows a banner saying it will retire on January 30, 2027. OpenAI’s migration guide still says Classic keeps getting model updates, bug fixes and security patches, and it didn’t mention a retirement date when we checked on November 3.
What OpenAI hasn’t said
The open questions are the ones that decide whether this is a cut. OpenAI hasn’t published the allowance for any plan in tokens, or said whether it is larger than today’s Work and Codex allowance to make room for chat. It hasn’t given a December start date, or said whether the five-hour and weekly windows stay as they are. Its ChatGPT release notes had no entry for the change as of November 3.
There is also nothing yet on plans outside Plus, Pro and Business. OpenAI’s article doesn’t address Free, Go, Enterprise or Edu, so people on those plans should assume nothing changes until OpenAI says otherwise.
OpenAI isn’t the only lab moving this way. TechCrunch notes that Anthropic merged Claude’s chat and Cowork screens in September. Anthropic has been reworking its caps too: its renamed Ultra plan has no five-hour limit but keeps a weekly cap.
The number to watch is the one in Settings, under Usage, once December arrives. If the combined bar looks about the size of today’s Codex bar, Plus users who lean on thinking-mode chat will notice fast. If OpenAI quietly enlarges it, the reassurance in its help article will turn out to be right.
Frequently asked questions
When do ChatGPT and Codex start sharing usage limits?
OpenAI says the shared allowance starts in December 2026 for Plus, Pro and Business plans. It hasn’t given an exact date. Once it begins, you’ll see your combined remaining usage and reset times in Settings, under Usage.
Does regular ChatGPT chat count against my Codex limit?
From December, yes, except for Instant chat, which OpenAI says won’t draw from your usage. How much each chat uses depends on the model and the task, and bigger models and more complex work use more. OpenAI hasn’t listed exactly which chat modes count.
Is ChatGPT Plus getting more expensive?
No price change has been announced. Plus is still $20 a month. What changes is how far the allowance goes, because chat and Codex will share it, and OpenAI hasn’t said how large the combined allowance will be.
What happens when I hit the shared limit?
You can wait for the limit to reset, upgrade your plan or buy credits to keep using eligible features. OpenAI’s Codex pricing page also says an agent that is already working can finish its current turn when you reach the limit.
Sources
- OpenAI Help Center, “Understanding ChatGPT and Codex usage limits” (updated about October 31 to November 1, 2026): help.openai.com
- OpenAI, Codex and ChatGPT Work pricing and usage limits (Learn page): developers.openai.com
- OpenAI Help Center, ChatGPT release notes (October 29, 2026 Pro usage entry; July 13 and 16, 2026 desktop app entries): help.openai.com
- OpenAI Help Center, ChatGPT Business release notes (Premium and Standard seats): help.openai.com
- OpenAI Help Center, “Moving to the new ChatGPT desktop app”: help.openai.com
- TechCrunch, “OpenAI is finally merging ChatGPT’s Chat and Work tabs” (Ivan Mehta, November 2, 2026): techcrunch.com
- 9to5Mac, “OpenAI is retiring ChatGPT Classic app, will soon combine Chat and Work tabs in new app” (Marcus Mendes, November 2, 2026): 9to5mac.com
- OpenAI, “Introducing ChatGPT Work, powered by Codex and GPT-5.6” (official YouTube channel, July 9, 2026): youtube.com
All sources accessed November 3, 2026. Prices and limits are as listed on OpenAI’s pages on that date.
Last updated: November 3, 2026. To report an error, see our corrections page. Articles are drafted with AI assistance and reviewed and edited by an editor; see our editorial policy.
