Sam Altman drops major GPT 5 update, unveils Auto, Fast and Thinking as response modes to choose; check rate limits here

After launching GPT-5 a few days ago, OpenAI CEO Sam Altman rolled out new updates today, 13 August, stating the users can now choose between three response modes, namely ‘Auto, Fast and Thinking’ and also fixed rate limits for ‘Thinking’ mode as 3,000 messages/week. For more capacity, he said GPT-5 Thinking mini can be undertaken.

According to him, most will use ‘Auto’; however believes extra control will benefit the users. With a new 196k-token context limit, the model can now process much longer documents and conversations. Sam mentioned that the limit may be revisited as usage patterns shift.

“You can now choose between “Auto”, “Fast”, and “Thinking” for GPT-5. Most users will want Auto, but the additional control will be useful for some people. Rate limits are now 3,000 messages/week with GPT-5 Thinking, and then extra capacity on GPT-5 Thinking mini after that limit. Context limit for GPT-5 Thinking is 196k tokens. We may have to update rate limits over time depending on usage,” Sam said.

GPT-4o comeback

GPT-4o has officially returned to the model picker for all paying users, Sam shared.

Sam shared that GPT-4o has been restored to the model picker for all paid users by default. He noted that if OpenAI ever decides to deprecate it, users will receive ample notice in advance, further mentioning that paid users now have access to a “Show additional models” toggle in the ChatGPT web settings, which allows them to add models like o3, 4.1, and GPT-5 Thinking mini. Sam further clarified that GPT-4.5 remains exclusive to Pro users, as it requires significantly more GPU resources.

“We are working on an update to GPT-5’s personality which should feel warmer than the current personality but not as annoying (to most users) as GPT-4o. However, one learning for us from the past few days is we really just need to get to a world with more per-user customization of model personality,” he added.

In its announcement last week, OpenAI stated that GPT-5 offers improved performance in coding and reasoning through complex problems. The company described it as advanced enough to elevate ChatGPT to the level of a PhD-level expert. It said, “GPT‑5 is our strongest coding model to date.”