Tibor Blaho says the ChatGPT web app now shows a hidden checkout disclaimer reading “Starting in December, plan usage will have a shared limit across all of ChatGPT and Codex.” His post attaches a single screenshot of the text. A Linux.do forum user says the same string appears in the checkout page’s front-end code.

Image shared by @btibor91 on X
Image shared by @btibor91 on X

Image credit: @btibor91 on X

What the checkout text appears to be

The forum user says the sentence is appended to the subscription terms paragraph on the Plus, Pro and Business checkout pages, before purchase, with a “Learn more” link to a help article. A second Linux.do post reproduces the string with a screenshot of the code.

The same forum post says the notice is gated by a flag called sharedUsageLimitEnabled, which defaults to false. If that is right, the disclaimer may only be visible to some checkout users for now.

How it compares with OpenAI’s current pages

OpenAI’s pages describe a split that the disclaimer would change. OpenAI’s Codex pricing page says ChatGPT Work and Codex share usage, and that Work usage inside ChatGPT uses the same pricing, credits and usage limits as Codex. A help article on GPT-6 Astra usage says Work and Codex share one allowance, while Chat’s message limits are separate and vary by plan.

The disclaimer’s wording goes further. It names “all of ChatGPT,” which would bring Chat into the shared pool.

A GitHub project post quotes official limits for GPT-6 Pro and GPT-5.6 Sol Pro that keep Chat, ChatGPT Work and Codex on separate allowances. That conflicts with the checkout text. The GitHub post is third-party, not an OpenAI page, so its wording cannot settle the question on its own.

What Tibo has said

Tibo, who leads communications for OpenAI’s Codex and ChatGPT, posted on October 28 that a team is working on an issue, that it is not rolling out yet, and apologised for any confusion. The text appears in the replies to Blaho’s post, and it is unclear which feature “this” refers to. In the same post, as quoted on that page, Tibo says the plan is to bring together the best of Chat and Work capabilities at some point this year.

Blaho suggests the disclaimer is tied to that work on combining Chat and Work. That is his inference, not an OpenAI statement.

A Tibo post shown on an X trending page reported that OpenAI is doubling usage limits for the Pro 100, 200 and 500 tiers and Business Premium. See our earlier report on the doubled Pro and Business Premium limits.

What to watch

The key checks are whether the December change appears in a help article, changelog or pricing page, which plans it covers, and whether OpenAI states the size of the shared pool and whether Chat messages draw from it. Until then, the Codex and Work sharing on current pages is the firmest ground.

Sources