AI & automation

ChatGPT Pro: allowance preserved, Chat and Work move closer

No announced allowance cut for Pro 200. A unified Chat and Work experience is planned, alongside shared usage for ChatGPT and Codex.

Generated symbolic model: one red power unit supplies two modules labelled ChatGPT and Codex.

OpenAI will not implement the announced reduction in the Pro 200 allowance. According to an email to subscribers available to our editorial team, the previous allowance for Work and Codex remains in place from 29 October. The company cites user feedback as the reason for reversing the change announced in its 29 September email.

The same message outlines a second development: Chat and Work are to come together in ChatGPT, with usage across ChatGPT and Codex counted against a shared allowance. These are related announcements, but they answer different questions: how much usage is included and how the tools are organised.

What stays the same for Pro 200

Subscribers addressed by the message keep their previous Work and Codex allowance. The planned reduction is cancelled. The email does not give a new fixed number of messages, tasks or hours and does not promise unlimited Work or Codex use.

That distinction matters for people using the tools for sustained projects. An unchanged allowance preserves the existing basis for planning. It does not mean that every task consumes the same amount or that all restrictions disappear.

From a conversation to a finished task in one place

OpenAI plans to test a unified Chat and Work experience with a small group over the coming weeks. Feedback from that group will inform a wider release. The announcement therefore does not mean that every account already has the new interface.

The intended workflow connects discussing an idea with carrying it out. Users will be able to bring in files and plugins, ask for a website or presentation and have tasks performed on the web. They should be able to review and steer the work as it progresses. Codex will remain a separate area in the ChatGPT desktop app, according to the email.

One allowance for ChatGPT and Codex

The message also announces a gradual move towards shared usage across plans. ChatGPT and Codex activity will count towards one allowance. Model choice and task complexity affect consumption: more advanced models and demanding tasks can use more of it.

For example, a brief editing request and a software change involving files, tools and checks need not have the same cost in usage. That is an illustration of the principle, not a conversion formula. OpenAI’s pricing and usage documentation also explains that task complexity affects consumption and that API prices are separate from subscription usage.

There is an explicit exception in the subscriber email: unlimited Instant use is included in the plan and will not count towards this allowance.

Checking usage and reaching the limit

Under Settings → Usage, users are to see their remaining total allowance and its reset time. If it is exhausted, the email lists three options: wait for the reset, upgrade the plan or purchase credits to continue using the relevant features. None of those options should be confused with a requirement to buy additional credits.

For day-to-day work, the account’s usage display is more useful than estimating capacity from the length of a prompt. Keep an eye on the remaining allowance before starting a long task and choose a model that fits what the task needs.

Our view: a smoother workflow still needs clear decisions

Bringing discussion and execution closer together can reduce handovers between tools. The real benefit depends on whether the resulting work is useful and reviewable. A website still needs suitable content, a presentation still needs a clear argument and software changes still need checks.

Our approach remains the same: define the result, provide the relevant context and review what the tools produce. We help businesses identify practical uses for AI and fit them into existing work. Talk to us about a concrete task.

Source and status of this article

This article is based on OpenAI’s email “Deine Pro-Nutzung und eine einfachere Art zu arbeiten”, supplied to our editorial team. It reports the announcement as of 1 November 2026. The individual allowance, availability and timing of changes should be checked in the relevant account. The cover image is a conceptual illustration and does not represent a measurable quota.

Let’s move forward

What would you like to change?

An idea, a specific challenge or your next step with AI. We’re listening.

Let’s talk