OpenAI is about to fold two usage buckets into one. Starting in December, subscribers on the ChatGPT Plus, Pro and Business plans will spend a single usage allowance on both ChatGPT and Codex, OpenAI's coding agent. The change sits in an updated Help Center article, "Understanding ChatGPT and Codex usage limits," stamped October 31, 2026.

The core sentence, quoted by several outlets that read the article on November 1: "Starting in December, Plus, Pro, and Business plans' usage allowance will be shared across ChatGPT and Codex." OpenAI also confirmed it by email to subscribers on November 1: "Your use of ChatGPT and Codex will draw from the same allowance."

What's confirmed, and what's exempt

Usage will be measured in tokens, which OpenAI describes as "the units of information a model processes and generates." More advanced models and complex tasks use more of the allowance. Combined remaining usage and reset times will show up in Settings, under Usage, and when you hit the cap you can wait for the reset, upgrade your plan, or buy credits.

The one named exception is Instant chat, OpenAI's fast-response mode. "Instant chat will not draw from your usage," the article says. Quick questions to the fast models stay free of the meter, which makes Instant the obvious pressure valve once the shared limit is live.

What OpenAI hasn't said

The gaps are the interesting part. There is no exact start date, just "December." There are no token figures for any plan, so nobody can yet say whether the single pool stretches as far as today's separate limits. OpenAI only offers a forecast: "We expect most users won't need to wait for their usage to reset more often than they do now." Forecasts aren't guarantees, and this one ships with no numbers behind it.

The article also stays silent on Free, Go, Enterprise and Edu plans, so don't assume the change reaches them, and on whether the five-hour windows on Plus and the weekly caps survive inside the combined system.

Our take

We checked the key lines against OpenAI's own Help Center article via multiple independent reads, and every figure of speech here is OpenAI's, not ours. The honest takeaway: this is a pricing detail that quietly reshapes how teams work. A coding agent grinding through a large repo can burn tokens far faster than chat, and from December it competes directly with every long conversation on your plan. Plan prices stay put at roughly $20 a month for Plus, $100 to $500 for Pro tiers, and about $20 per user for Business on annual billing. The smart move right now is boring but real: open Settings, then Usage, and learn what your current burn looks like before December merges the meters.