OpenAI merges ChatGPT and Codex limits: every chat now eats your budget

OpenAI is rewriting the rules for everyone who pays for ChatGPT. Starting in December, Plus, Pro and Business subscribers will get a single shared allowance across ChatGPT and Codex. The company spelled it out today in its help center.

The allowance will be measured in tokens, the chunks of text a model reads and writes. How much a request costs depends on the model you pick and how hard the task is: stronger models and heavier jobs burn through it faster. There is one exception: the quick Instant mode won't draw from it at all.

OpenAI's reasoning is that it is moving the agentic features of ChatGPT Work into the main ChatGPT experience. Today Work, Codex and several agent features already share one pool, while regular chat is counted separately. In December that wall comes down.

Rumors had been circulating for a few days, after a line about a shared limit starting in December turned up in the ChatGPT checkout page code, but the company hadn't confirmed anything. Now it is an official help page.

The company promises most users won't hit a reset more often than they do now. Remaining usage and reset times will show up under Settings, Usage. Run out and you can wait, upgrade, or buy credits. OpenAI hasn't said how many tokens each plan will get.

If you live in Codex all day, pay attention: every long chat will now take a bite out of the same pie. We'll see in December whether "most users will be fine" holds up.