The Next Chapter for GPT-5: OpenAI Addresses User Feedback with New Features

OpenAI has announced a significant upgrade for GPT-5, introducing multiple thinking modes and increasing rate limits for ChatGPT Plus subscribers. This update comes just a week after the model’s launch, which was met with widespread criticism from the user community.

New Thinking Modes for GPT-5

To address user backlash over the lack of choice, OpenAI CEO Sam Altman confirmed on X that GPT-5 now offers several thinking modes: Auto, Fast, Thinking, and Thinking-mini. These modes give users more control over how the AI processes requests.

• Auto: This mode lets GPT-5 decide the optimal amount of time to spend thinking.
• Fast: For instant responses, this mode provides quick answers without extensive processing.
• Thinking-mini: This option offers a balance, thinking quickly to provide good answers.
• Thinking: This mode takes longer to process, aiming to deliver higher-quality, more detailed responses.

This change marks a U-turn from OpenAI’s earlier decision to limit user choice and remove access to legacy models. The company recently made GPT-4o available again for paid subscribers, and these new thinking modes further demonstrate a commitment to user-driven feedback.

Increased Rate Limits for ChatGPT Plus Subscribers

In another major update, Altman announced a significant increase in the rate limits for ChatGPT Plus subscribers. The initial limit of 200 messages per week for the GPT-5 “Thinking” model has been raised to 3,000 messages a week. This was a key point of discontent for paid users who felt the initial limits were too restrictive for the premium tier.

Altman also noted that the context limit for GPT-5 Thinking is 196k tokens and that rate limits may be adjusted over time based on usage. While earlier hints of a potential GPT-5 Pro for Plus subscribers have been walked back due to “compute” limitations, these updates show OpenAI is actively working to rebuild trust with its user base.

With these new features and increased access, OpenAI is taking steps to right the initial wrongs of the GPT-5 launch and give users the performance and flexibility they expect.