OpenAI CEO Sam Altman announced a major update to ChatGPT. This update introduces different modes that users can choose from for the GPT-5, including “Auto”, “Fast”, and “Thinking”.
hackbar notes that Altman has tweeted that users can now choose between different modes of GPT-5 on social media. Most users will probably choose the “automatic” mode, but the “think” mode will be very useful for users who need extra control.
The limit for GPT-5 Thinking is currently 3000 messages per week, and after this limit is exceeded, GPT-5 Thinking mini will provide additional capacity. The context limit for GPT-5 Thinking is 196,000 tokens. OpenAI may update the limits at any time based on usage.
Additionally, the 4o model is now re-added to the model selector by default for all paying users. If the decision to deprecate the model is ultimately made, OpenAI will give users advance notice. Paid users can now also switch “Show other models” in ChatGPT web settings to add models like o3, 4.1, and GPT-5 Thinking mini. Note that the 4.5 version is only available to Pro users as it requires a significant amount of GPU resources.
Altman also said that OpenAI is updating the personality of GPT-5 to be warmer than the current personality, but not as annoying as GPT-4o. Experience over the past few days has shown that OpenAI needs to customize the model personality more according to user needs.