People splitting work between ChatGPT and Codex will have a shared allowance to watch from December. OpenAI’s help page says the change applies to Plus, Pro and Business as Work’s agentic capabilities move into the main ChatGPT experience.
The task matters more than a simple request count
Usage will be measured in tokens, the information units a model processes and generates. Two requests need not consume the same amount. The selected model and task complexity affect usage; more advanced models and complex work generally draw more. Instant chat will not draw from the allowance.
A combined view of what remains
Settings → Usage will show the remaining combined allowance and reset times. At a limit, the documented options are to wait for a reset, upgrade the plan or buy credits for eligible features. This is a transition announced for December, rather than a claim that every account already uses the new arrangement.
Prepare a workload, not a guessed price list
Separating short information requests from longer coding assignments can make a work plan easier to follow. It does not provide an exact token cost in advance. The useful feedback will come from the account’s real usage display, which can inform later choices about the model and the task.
Milo’s assessment
Start with the job you need done. A shared allowance adds another reason to define the required result before choosing a model. Plan around observed capacity rather than treating every message as an identical unit of work. That makes the new combined view useful for decisions, not simply another number to watch.
