OpenAI has doubled the usage limits for several of its professional and business subscription tiers, reversing a previous plan to significantly reduce capacity for its high-end users. The changes, effective October 29, 2026, impact the ChatGPT Pro 100, Pro 500, and Business Premium plans.
The move also includes the cancellation of a planned 50% usage cut for the $200-per-month Pro 200 plan. OpenAI Head of Product Tibo announced the updates as part of the company’s “28 Days of Shipping” campaign. The reversal follows an earlier attempt to manage infrastructure strain caused by the launch of the GPT-6 Astra model in September, which had led to unprecedented demand on OpenAI’s processing capacity.
Restoration of Pro 200 Limits
For users on the legacy Pro 200 tier, the announcement serves more as a restoration of service than a new increase. OpenAI had originally scheduled a 50% reduction in usage limits to take effect this week.
With the cut now canceled, these users retain their original capacity. While the Pro 100 and Pro 500 tiers are seeing their active usage limits for ChatGPT Pro effectively double from their previous baseline, the Pro 200 tier remains at its established level.
Understanding Work, Codex, and Dots Allowances
The increased limits specifically apply to “ChatGPT Work” and “Codex” features, which share a combined usage allowance. These tools are distinct from standard ChatGPT message caps and are designed for developers and enterprise users who require high-frequency interaction with the models for coding and complex automation.
OpenAI also clarified the usage status of “Dots,” the company’s recently introduced “always-on” AI assistants. Currently, interactions with Dots do not draw from the primary Work and Codex allowances. This separation allows users to maintain background AI agents without depleting the quotas needed for active development or content generation tasks.
The decision to expand capacity comes as OpenAI faces increased competition in the professional AI space. The expansion of these limits provides a broader buffer for power users who rely on the platform for production environments, signaling a stabilization in the company’s compute resources following the Astra model rollout.
