What happened
OpenAI confirmed it will merge the token allowances for ChatGPT and Codex into a single unified quota beginning December 2026, according to a CryptoBriefing report Friday. Until now, paying users have operated against two distinct meters: one tied to ChatGPT conversational usage, and a separate allowance governing Codex, the company's agentic coding environment that handles longer-running developer tasks.
The two surfaces were billed and throttled independently. Under the new structure, both will draw from the same pool. The company presented the change as a budgeting and transparency step for subscribers who have struggled to track consumption across the two products.
No revised pricing was published alongside the announcement, and OpenAI did not confirm whether existing tier ceilings will be raised to accommodate the combined load. CryptoBriefing's report is the first write-up to carry the detail.
Why it matters
The split allowance has, until now, functioned as a soft subsidy for developers. Codex users running agent loops that chew through tokens on long autonomous coding sessions did not have those draws counted against their ChatGPT quota, and vice versa. Consolidating the two under one ceiling means a heavy Codex day now eats into the same budget as a heavy chat day.
For a Plus or Pro subscriber who leans on both, that is a real behavioral change. The alternative read is simpler. One meter is easier to understand than two, and OpenAI has been pushing toward a more legible pricing surface for twelve months.
The headline is a cleanup. The flow picture is competition for the same pool. Which read wins depends entirely on where the new combined cap lands relative to the sum of the old two, and that number was not in the announcement.