BleepingComputer reported on 29 October that OpenAI dropped a planned usage cut for existing Pro 200 subscribers, who pay $200 a month. An OpenAI staffer known as Tibo said on X that the change is live and covers ChatGPT Work and Codex, which share one allowance. OpenAI's help page says the cut, announced in a 29 September email, is canceled.
Tibo said limits on Pro 100, 200 and 500 and Business Premium doubled. OpenAI's help page gives detail: Pro 100 and Pro 500 users now get twice the Work and Codex allowance. Pro 200 users who already had the higher allowance keep it. Those on the lower one now match it. That is why Tibo also said Pro 200 usage stays the same. OpenAI gives no exact numbers per tier.
An allowance is a metered budget, not a count of tasks. OpenAI says use depends on the model, context size, reasoning and tool use. Fast mode spends the allowance at 2.5 times the Standard rate. Ultrafast spends it at 8 times, but only on Pro 500 and only for GPT-6 Astra and GPT-6.1 Sol. These rates show how fast the allowance drains, not promised speed. So a doubled limit may not buy twice the finished work. Buyers should judge plans by how heavy users work.