The late-September notice that had subscribers up in arms was ultimately reversed by OpenAI itself. On October 29, the company announced an immediate increase to the usage limits for Work and Codex in ChatGPT Pro and Business Premium—the Pro 100, Pro 200, and Pro 500 tiers all doubled outright. New users get it right away, and existing users see it take effect automatically without waiting for renewal; expanded usage for Dots, the always-on AI assistant, is calculated separately. This reversal came just one day before the date it had originally planned to "make cuts."

The controversy began on September 29. At the time, OpenAI, citing "improved model efficiency," notified users that it planned to cut Pro 200 limits by about half and move existing users to a lower tier after October 29. Subscribers immediately erupted: the monthly fee wasn’t going down, but their benefits were being halved—this amounted to breaking a promise to paying customers. Curiously, if efficiency was improving and costs were falling, why cut user benefits at all? Amid a strong backlash, OpenAI ultimately reversed course and made clear it would not lower the limits for Pro 100, Pro 200, Pro 500, or Business Premium.

But the company also threw cold water on the idea: doubling the quota does not mean you can actually complete twice as many tasks. Actual output depends on the model, task complexity, context length, tool calls, and speed mode. In particular, Fast mode consumes about 2.5 times the quota of Standard mode, while Ultrafast goes all the way up to 8 times—if you run hard in high-speed modes, even the doubled quota may quickly run dry. For heavy users, this "backing down" looks more like psychological reassurance than a real boost in capacity.