OpenAI to give Plus, Pro and Business users a single shared token budget across ChatGPT and Codex from December, as Chat and Work merge

According to an updated OpenAI support page (as reported by The Decoder, which we have not seen directly), from December the Plus, Pro and Business plans will share one token-based usage budget across ChatGPT and Codex, replacing separate limits. Users will see their balance and reset times in Settings. At the limit they can wait for a reset, upgrade or buy extra credits. Chats with Instant models will not count toward the budget. No token amounts are given, so the size of the budget, and whether it is smaller or larger than today's limits, is unknown.

Paying users who lean heavily on Codex or the more capable models are the most exposed. A shared pool means heavy use in one product now eats into the other, and buying credits becomes a more visible route to extra capacity. Teams on Business plans should check how their usage splits between chat and coding before December, because budgeting gets harder when limits are opaque. For light users the change is probably small, and the Instant exemption suggests everyday chat is largely unaffected. Until the actual numbers appear, the real impact is uncertain.

The headline claim that this is "the end of all-you-can-prompt pricing" is The Decoder's interpretation, not something OpenAI said, and limits on paid plans have existed for a long time. The predictions of lower-quality results for heavy users and the explanation that the merge answers complaints about complexity are speculation. The Chat and Work merger itself was already announced, so only the shared budget is new.

1. 01The actual token allowances per plan once OpenAI publishes them, compared with current ChatGPT and Codex limits, to show whether heavy users really get less.
2. 02User reports in December on hitting limits early, plus any change to credit prices or plan prices, which would confirm or refute the 'end of unlimited' reading.
3. 03Whether Anthropic, Google or others move to similar shared or token-metered budgets on consumer and team plans, which would show this is a trend rather than one company's adjustment.