ChatGPT usage limits are about to work in a new way. OpenAI has told users it will merge the Chat and Work tabs in ChatGPT. Starting in December, ChatGPT and Codex will also draw from one shared token allowance on Plus, Pro and Business plans.

I'm Nitin, and I coordinate projects for the dev teams at Debugged Pro. Our developers use Codex for code, and our project managers use ChatGPT for docs and research. So a single pool matters to us, and probably to you too. Here are the verified facts and a simple plan for agencies and businesses in India and the USA.

Illustration of ChatGPT and Codex sharing one usage allowance — Image credit: The New Stack, from its report on OpenAI's shared ChatGPT and Codex usage limit.

What OpenAI Announced

The news first came through an email OpenAI sent to users. TechCrunch reported on the Chat and Work tab merge on November 2, 2026. Right now, the ChatGPT desktop app has ChatGPT and Codex sections. Inside ChatGPT, there are separate Chat and Work tabs.

OpenAI plans to fold these into one place. In its email, the company said the change will let you "move from exploring an idea to having ChatGPT carry it through, all in one place." It added that you can bring in files and plugins and ask ChatGPT to build a website, create a presentation or do things on the web.

Here are the key points:

Tabs: Chat and Work will become one experience inside ChatGPT.
Codex: It stays as a separate section in the desktop app.
Rollout: Select users get it first for feedback, then a wider release.
Usage: ChatGPT and Codex will draw tokens from the same allowance pool.
Timing for limits: The shared allowance starts in December for Plus, Pro and Business plans.

OpenAI has not shared a date for the full tab merge. It has also not said how many tokens each plan will get.

How the New ChatGPT Usage Limits Will Work

OpenAI explains the details in its help article on ChatGPT and Codex usage limits. Usage will be counted in tokens. Tokens are the small units of text a model reads and writes.

How fast you use your allowance depends on two things:

The model you pick. More advanced models use more of your allowance.
The task you give it. Long, complex jobs use more than short questions.

There is one useful exception. Instant chat will not draw from your usage. So quick everyday questions on the fast mode stay outside the pool.

You will be able to check your combined remaining usage and reset times in Settings, under Usage. If you hit the limit, OpenAI lists three options. You can wait for the reset, upgrade your plan or buy credits for eligible features.

What is shared today and what changes

This is not a total change from scratch. According to The New Stack's report on the shared usage limit, Codex, ChatGPT Work, ChatGPT for Excel and Workspace Agents already share one agentic allowance. Regular ChatGPT conversations have their own separate limits.

That separation is what ends in December. After the switch, a long Codex session can eat into the allowance you would use for normal ChatGPT chats. The reverse is also true.

Area	Today	From December
Regular ChatGPT chats	Separate limits	Shared token pool
Codex, Work, Excel, Workspace Agents	Shared agentic allowance	Same shared token pool
Where to check	Varies by product	Settings, Usage page

Why Some Users Are Unhappy

OpenAI says it expects most users won't need to wait for resets more often than they do now. TechCrunch noted the company's own warning that some people may need to wait for resets more frequently, depending on usage and model.

Not everyone is convinced. The New Stack points to a thread on OpenAI's developer forum where users asked the company to keep the limits separate. One user wrote, "ChatGPT and Codex serve different purposes, and their usage limits should remain separate." Another asked for a real-time dashboard showing usage by product, model and task before the change goes live.

The New Stack also asked OpenAI how big the shared pool would be. The company declined to answer and pointed to its existing documentation.

The Pro plan reversal from October

The timing adds to the doubt. On October 29, OpenAI reversed a plan to cut Work and Codex usage for its $200 Pro tier. It also doubled allowances for the Pro 100, Pro 500 and Business Premium tiers, as reported by The New Stack.

One forum user summed up the worry well. Doubling an allowance while adding more things that use it does not always mean more usable capacity. That is a fair point for any team that already runs close to its weekly Codex limit.

Image credit: Samuel Boivin/NurPhoto via Getty Images, as used by TechCrunch.

What ChatGPT Usage Limits Mean for Agencies

For a solo user, this is mostly about convenience. For an agency, it is about delivery. If a developer runs out of tokens on a Friday evening before a client release, that is a real problem.

Here is how the change can show up in daily agency work:

Mixed roles on one seat: If a tech lead uses ChatGPT for client emails and Codex for code reviews, both now drain one pool.
Agent tasks cost more: Asking ChatGPT to build a deck or a website is a heavy task. It will use more tokens than a quick question.
Model choice matters more: Picking the top model for every small job will burn the allowance faster.
Budget surprises: Buying credits is an option, but it adds a cost line many agencies don't track today.

Teams in India often work late into the US day to overlap with clients. That means heavy usage can cluster in the same hours. A shared pool makes it more likely that a seat hits its cap right when the client is online.

A quick example

Say your agency has a five-person team on Business seats. Two developers use Codex for refactoring all day. Three project staff use ChatGPT to draft specs and summarise client calls.

Today, the developers' Codex work and the project staff's normal chats sit on separate limits. From December, each person's chats and Codex runs come out of one allowance. The developers who also use ChatGPT for research may feel the pinch first.

A 6-Step Plan to Manage ChatGPT Usage Limits

You don't need to panic or switch tools. But a little planning now will save stress in December. Here is the simple plan we are following at Debugged Pro.

Map who uses what. List each seat and note whether that person mostly chats, uses Work, or codes in Codex.
Check Settings, Usage weekly. Once the new view is live, note how fast each seat drains. Write it down for two or three weeks.
Use Instant chat for quick questions. It does not draw from the pool, so train your team to use it for simple lookups.
Match the model to the job. Save advanced models for complex code and long tasks. Use lighter options for drafts and summaries.
Protect release days. Before a big client delivery, avoid running large agent tasks on the same seats that handle the release.
Set a credit budget. Decide in advance how much your agency will spend on extra credits each month, and who can approve it.

If your team builds custom tools on OpenAI's API instead of the ChatGPT app, these plan limits work differently. Our ChatGPT development services page explains how we set up API-based tools with their own budgets.

How This Compares With Other AI Tools

OpenAI is not the only company joining chat and agent work. TechCrunch noted that Anthropic merged Claude's chat and Cowork functions into one screen in September.

Usage caps are tightening across the market too. We recently covered how Grok Bot usage limits and AI agent costs changed for teams. The pattern is clear. AI vendors want one simple app, and they want to meter heavy agent work more closely.

For agencies, the lesson is to avoid building every workflow around one vendor's plan limits. Keep a backup tool ready for critical tasks. That way, a reset timer never decides your delivery date.

Should Your Team Change Plans Now?

Not yet. OpenAI has not published the size of the new pool for any plan. Upgrading now, before you see real numbers, means paying for capacity you may not need.

A better move is to watch your usage for the first few weeks of December. If a seat hits the cap often, then compare the cost of an upgrade with the cost of credits. If you need help planning tools and budgets for a larger project, our software development consulting team can review your setup.

Key Takeaways

ChatGPT usage limits will be shared with Codex from December on Plus, Pro and Business plans.
Usage is counted in tokens. Bigger models and complex tasks use more.
Instant chat does not draw from the shared allowance.
The Chat and Work tab merge reaches select users first. Codex stays a separate section.
OpenAI has not shared token amounts per plan, so track usage before you upgrade.

Final Thoughts

The merged tab should make ChatGPT easier to use. The shared pool is the part agencies need to watch. Most teams may see no change, but heavy Codex users could hit resets sooner.

Track your usage, match models to tasks, and keep a credit budget ready. With those habits in place, December's change should be a small adjustment, not a delivery risk.

About the Author

Nitin Kumar is the Lead Project Coordinator at Debugged Pro. He manages delivery for web, mobile and AI projects for clients in India and the USA. He writes about practical ways teams can use AI tools without losing control of time, cost or quality.