Did the $100 Plan Affect the GPT-5.4 Pro Model?

A significant controversy is brewing among power users of OpenAI's ChatGPT Pro service. Following the introduction of a new, cheaper $100 per month subscription tier, users who downgraded from the original $200 plan are reporting a drastic reduction in the performance of the GPT-4 Pro model. Specifically, they claim that complex, long-running queries that previously took 30 to 60 minutes to process now time out or complete in only 10 to 15 minutes. This perceived throttling has been described as a 'devastating change' and a 'massive downgrade' for professional workflows that depend on the model's deep reasoning capabilities for core planning and design tasks.

The user reports suggest OpenAI may have implemented performance limits on the GPT-4 Pro model for the lower-cost tier, a move critics are calling a 'stealth nerf.' This development raises critical questions about transparency in AI service pricing and feature segmentation. For professionals who integrated the Pro model's extended processing as a central anchor in their work—often chaining it with tools like Codex—the reduced runtime effectively cuts the AI's 'thinking' capacity by more than half. The situation highlights the growing tension between making advanced AI accessible and maintaining the uncompromised performance that high-paying enterprise and professional users require.