OpenAI Reintroduces Older Models Amid User Feedback on GPT-5
OpenAI has reversed its all-in approach with GPT-5, reintroducing older models like GPT-4o, GPT-4.1, and o3 for paid users. The update allows subscribers to toggle between speed-first, deep-think, or auto-select modes, addressing complaints about tone and accuracy. CEO Sam Altman framed the shift as offering more control, with rate limits now set at 3,000 messages/week for GPT-5's 'Thinking' mode.
The MOVE transforms ChatGPT into a modular toolkit, where users match specialized AI models to specific tasks. While GPT-5 was marketed as a unified solution, its limitations in handling niche requirements prompted this strategic retreat. The decision highlights the tension between generalized AI capabilities and domain-specific performance demands.
