ChatGPT and Codex will draw from a single usage quota starting in December

On October 29, OpenAI reversed course. The company had planned to cut the usage included in the $200-a-month Pro plan after that date, without lowering the price. Faced with subscriber anger, it ended up doubling the limits. “We are doubling all usage limits for all Pro 100/200/500 and Business Premium plans, effective today,” Thibault Sottiaux, described by BleepingComputer as the face of ChatGPT and Codex on X, wrote on X. Shortly afterward, an email to subscribers announced a deeper change.

According to that email, cited by BleepingComputer on November 1, ChatGPT usage and Codex usage, OpenAI's programming assistant, will draw from the same allowance. According to TokenPost, the change will take effect in December, with no specific date, and will apply to Plus, Pro and Business subscribers. The allowance will be measured in tokens (tokens, the units of text that models process).

What will count, and what won't

The simplest mode, called “Instant,” will remain unlimited and will not be counted. However, the most advanced models and complex tasks will consume more, though OpenAI does not say how much. The company also does not publish the size of the allowance for each plan. It says it expects most users will not wait for their quota to reset more often than they do today, a forecast that cannot yet be verified.

The remaining balance and the reset time will appear in Settings, under Usage. Once the allowance runs out, users will have to wait, move to a higher plan, or buy credits for certain features.

Why merge them?

OpenAI ties this change to another one: bringing the features of ChatGPT Work, its mode that carries out tasks for the user, into the standard conversation. In the coming weeks, a small group of testers will be able to move from a simple chat to a heavier task without switching modes: creating a website or a presentation, or acting on the web while the user checks the work. “This will let you go from exploring an idea to having ChatGPT carry it out, all in the same place,” OpenAI promises in its email. Codex will remain a separate space in the desktop app.

What businesses should watch

For a team that codes heavily with Codex and writes with ChatGPT, a single allowance could run out faster than two separate ones. That can only be measured once the volumes are published. Until then, two habits are worth adopting: note your actual consumption of both tools now, and budget for the possible purchase of credits.

Several points remain open. Plus and Business subscribers currently have a limit per five-hour window, which Pro subscribers do not, and how that fits with the shared allowance has not been explained. OpenAI has also given no details on the timeline in Europe. In any case, the October 29 episode shows that these rules can still change.