OpenAI says that starting in December, ChatGPT and its Codex coding tool will draw from one shared usage allowance for Plus, Pro and Business subscribers. Usage will be measured in tokens. The announcement leaves out the size of the allowance, so the practical impact is not yet clear.
Contents
show
What OpenAI has said
The change is described in an OpenAI Help Center article updated on November 2, as summarised by Notebookcheck:
- Timing: “in December”, with no specific date.
- Plans: Plus, Pro and Business. Free and Go plans are not mentioned.
- Measurement: tokens. Heavier models and more complex tasks generally use more.
- Exclusion: regular chats with Instant will not draw from the allowance.
- Not stated: the size of the shared allowance on any plan, and how it interacts with existing per-plan limits.
Notebookcheck also reports that ChatGPT will show combined remaining usage and reset times under Settings, then Usage. That is the outlet’s account rather than an OpenAI statement.
Who this affects
- Chat-only users who rarely use Codex: probably little change, since Instant chats are excluded.
- Heavy Codex users: coding sessions may now compete with chat for the same pool, so a long agentic run could reduce what is left for everything else.
- Business teams: admins should watch for plan documentation on how a shared pool is split between users.
What to do before December
- Note how often you use Codex today and which model you select, since heavier models use more tokens.
- Re-check OpenAI’s Help Center article for the allowance size once it is published; do not assume it matches your current limits.
- Compare plan prices with alternatives such as Claude Pro and Gemini Advanced only after the numbers are public.
For a broader comparison of the assistants, see ChatGPT vs Claude vs Gemini.
