TL;DR: OpenAI has updated its Help Center to say that, starting in December, ChatGPT and Codex usage limits on the Plus, Pro and Business plans will come from one shared allowance. Usage will be measured in tokens, so bigger models and harder tasks will burn through the allowance faster, while quick Instant chats won’t count at all. OpenAI says it expects most people won’t hit resets more often than they do today, but it has not published token totals per plan or an exact start date. You will be able to track the combined balance in Settings → Usage.
If you pay for ChatGPT and also use Codex for coding, the way your ChatGPT and Codex usage limits work is about to change. According to a newly updated OpenAI Help Center article, Plus, Pro and Business subscribers will soon draw on a single pool for both products instead of separate buckets. Here is what OpenAI has said, what it hasn’t, and what heavy users should do before December.
What OpenAI announced
The Help Center page, titled “Understanding ChatGPT and Codex usage limits,” was updated recently, and the change was picked up by AI newsletters on November 1. Its key line: “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.”
OpenAI lists four main points on the page:
- One shared allowance: ChatGPT and Codex usage will draw from the same pool on Plus, Pro and Business.
- Measured in tokens: usage will be counted in tokens, the units of text a model reads and writes. How much you use depends on the model and the task.
- Instant chat is free of the meter: OpenAI says “Instant chat will not draw from your usage.”
- A single dashboard: combined remaining usage and reset times will appear in Settings → Usage. If you run out, you can wait for the reset, upgrade or buy credits.
OpenAI also says it is “preparing carefully for this transition” and that it expects “most users won’t need to wait for their usage to reset more often than they do now.” That is the company’s own expectation, not an independently measured result.
Why the change is happening now
The shift is tied to ChatGPT Work, OpenAI’s agent mode for longer, multi-step jobs such as research reports, spreadsheets and presentations. OpenAI’s ChatGPT Work and Codex help page already says “Work follows the same usage structure as Codex,” and the Codex plan page notes that Codex, ChatGPT Work, ChatGPT for Excel and Workspace Agents share an allowance and credit pool where those features are available.
In other words, agentic work in ChatGPT and coding in Codex were already metered together. December’s change folds that metering into the main ChatGPT experience as Work’s agent features move there, so most of what you do beyond quick Instant answers will count against one balance.
Old vs new: how usage limits compare
| Today | From December (per OpenAI) | |
|---|---|---|
| Regular Chat | Separate model and message limits | Instant chat does not draw from the shared allowance |
| Work and Codex | Already share an allowance | Share one allowance with the rest of ChatGPT’s agentic features |
| How usage is counted | Estimated messages per window, varying by task | Tokens, varying by model and task |
| Where to check | Usage dashboard / /status in Codex CLI | Settings → Usage, showing a combined balance |
| If you run out | Wait, upgrade or buy credits | Wait, upgrade or buy credits |
How much usage do you get?
OpenAI has not published the size of the new token allowance for any plan. The closest guide is the current Codex pricing documentation, which gives rough local messages per five-hour window for Plus and Standard Business seats. OpenAI stresses these are estimates, not fixed limits:
| Model | Plus / Standard Business (local messages per 5 hours) |
|---|---|
| GPT-6 Astra | 5–45 |
| GPT-6.1 Sol | 15–160 |
| GPT-6.1 Terra | 35–285 |
| GPT-6 Sol | 15–150 |
| GPT-6 Luna | 350–3,000 |
Source: OpenAI’s Codex pricing docs. According to the same page, Pro plans currently have no five-hour limit but weekly limits may apply, and cloud tasks may use more allowance than local messages. The spread in each row shows why tokens matter: a short script and an hour-long refactor are both “one message,” but they cost very different amounts.
What eats your allowance fastest
OpenAI’s documentation names the factors that drive usage: the model you pick, task complexity, context length, reasoning effort, tool use and where the task runs. Practically, that means:
- Frontier models cost more. GPT-6 Astra uses the allowance faster than lighter models, according to OpenAI. Our explainers on GPT-6.1 Terra and GPT-6 Sol vs Luna cover where the cheaper models fit.
- Long context adds up. Large repositories, long chats and big file uploads all add tokens to every turn. See what a context window is for why.
- Long-running agent tasks are the big spenders. OpenAI says a long task “can use substantially more than a short request.”
What OpenAI hasn’t said yet
The Help Center page is short, and several details are still missing:
- No exact December start date.
- No token figures for Plus, Pro or Business.
- No word on whether five-hour and weekly windows stay the same under the shared pool.
- No changes listed for Free, Go, Enterprise or Edu plans, which the page doesn’t cover.
As the DailyAI Briefing put it, heavy Codex users and heavy chatters will now be spending from the same pot. Someone who codes all afternoon may find less room left for an evening Work research task, and vice versa.
What to do before December
- Check your current usage pattern. Open the usage dashboard (or run
/statusin Codex CLI) for a week or two to see how close you get to your limits today. - Match the model to the job. Use lighter models for routine edits and save the frontier model for hard problems.
- Use Instant for quick questions. Since OpenAI says Instant chat won’t draw from the allowance, simple lookups there won’t eat into your Codex time.
- Know your fallbacks. Plus and Pro users can buy credits, and OpenAI’s docs note you can run extra local Codex chats with an API key at standard API rates.
- Business admins: review seat types. OpenAI says Premium seats get a higher included allowance for Work and Codex than Standard seats.
If you’re still deciding whether a paid plan is worth it, our breakdown of whether ChatGPT Plus is worth it is a good next read.
Bottom line
From December, OpenAI plans to merge ChatGPT and Codex usage limits on Plus, Pro and Business into one token-based allowance, with Instant chat left off the meter. OpenAI expects little disruption for most users, but until it publishes real numbers, the safest move is to learn your own usage now and lean on cheaper models for everyday tasks. We’ll update this article when OpenAI shares plan-level figures.
The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.
