What happened
OpenAI employee Tibo Sottiaux announced on X that the company is doubling usage limits for all Pro 100, Pro 200 and Pro 500 tiers, as well as Business Premium, effective immediately. The statement follows a planned reduction in included usage for existing $200 Pro subscribers after October 29, with no change to the subscription price. OpenAI’s announcement says the reduction will not happen; the limits for the named plans will instead increase.
The practical effect is most relevant to ChatGPT Work and Codex, which share usage allowances. A higher allowance for the account therefore affects capacity across both products rather than creating an independent quota for each. Users who switch between work-oriented ChatGPT features and Codex should understand the change as an increase to their shared pool, not necessarily a separate doubling for each interface.
The announcement also distinguishes Dots extended limits from default usage: those limits remain separate. It names the affected subscription tiers but does not give exact numerical limits for any of them. The immediate effective date means subscribers do not need to wait for a future rollout date, though the public statement does not describe how users can inspect their account-specific allowance or how the doubling is calculated across different usage patterns.
The sequence is notable for teams planning around a paid product’s operating ceiling. OpenAI had been expected to reduce included usage for existing $200 subscribers after October 29, while maintaining the price. The new statement replaces that planned reduction with a broader increase spanning three Pro price points and Business Premium. It does not say that every feature, model or workload has an identical quota, so the doubling should not be interpreted as a blanket promise of twice as many actions in every workflow.
What the archive adds
New Runtime’s earlier coverage reported that OpenAI planned to halve Work and Codex usage for the $200 Pro tier from October 30: the October 24 note. A separate October 31 note covered the planned shared allowance across ChatGPT and Codex. This is a new fact on that known story: OpenAI has reversed the reduction and says it is doubling limits across additional paid tiers.
Trend signal
This weakens the expectation that inference scarcity must always translate into tighter customer-facing caps: OpenAI has reversed a planned cut and announced higher limits. But it does not establish that limits are disappearing or that capacity constraints have eased across the market. The relevant assumption is tracked in Compute-backed AI contracts; this change is evidence against its tightening-only direction, not a full contradiction of the broader capacity thesis.
Why now
The change follows OpenAI’s planned reduction of included usage for existing $200 Pro subscribers after October 29. Sottiaux’s statement makes the reversal effective immediately, replacing the expected cut with a higher allowance for the named tiers.
Why it matters
Usage limits are part of the effective price of a subscription. A nominally unchanged monthly fee can buy substantially different practical capacity when a provider changes what subscribers may do with shared allowances. For teams using ChatGPT Work and Codex, the announced doubling may reduce the need to ration usage or move work between accounts, but the lack of published numerical quotas prevents a reliable estimate of how much additional work each team can complete.
The episode is also a reminder not to treat a planned quota change as settled until the provider confirms it is in force. Teams that build internal cost forecasts or usage policies around a tier should record the stated allowance separately from the headline price, and verify account-level limits after a change. Because Work and Codex share allowances, monitoring only one surface could obscure where the capacity is being consumed. The announcement provides useful direction, but not enough detail to translate it into a fixed number of additional tasks.
Limits and uncertainty
OpenAI’s public statement does not publish the exact numerical limits for Pro 100, Pro 200, Pro 500 or Business Premium, nor explain whether the doubling is measured per model, feature, time window or another quota unit. The announcement identifies ChatGPT Work and Codex as the main affected products and says Dots extended limits remain separate from default usage, but it does not provide account-level examples. The change has not been independently measured, and users’ realized capacity may depend on product-specific limits not described in the statement.
Reader takeaway
OpenAI has replaced a planned Pro usage cut with an immediate doubling of limits across its three named Pro tiers and Business Premium. Builders using Work and Codex may have more shared capacity, but should verify actual account limits rather than budgeting from an assumed number: OpenAI has not published the new quota values.
Open question
What are the new account-level quotas for each tier, and how are they divided across ChatGPT Work and Codex?