OpenAI to Improve ChatGPT 5’s Personality After User Backlash, GPT-4o Makes a Comeback

OpenAI is improving the personality of its flagship ChatGPT 5 model after facing criticism that the newer model feels colder in responses compared to GPT-4o. OpenAI CEO, Sam Altman, shared a post on X explaining the new changes, “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.”

After releasing GPT-5, OpenAI faced a wave of backlash as the company abruptly deprecated and replaced GPT-4o with GPT-5 as the default model on ChatGPT without prior notice. Many long-time users who loved interacting with the GPT-4o model were caught off guard and some even threatened to cancel ChatGPT subscription if GPT-4o was not restored.

While OpenAI had worked hard to reduce the sycophantic, people-pleasing behavior of GPT-4o, users ultimately loved that model’s personality. In response to the outcry, OpenAI finally restored the GPT-4o model, but only for ChatGPT Plus and Pro users. Along with that, Altman now says that GPT-5 is also being improved to make it more conversational and engaging.

Apart from these changes, there is now a new interface that lets you choose the reasoning effort for the GPT-5 model. ChatGPT Plus users can choose between “Auto”, “Fast”, and “Thinking” modes, depending on the complexity of the user query.

Altman writes, “Most users will want Auto, but the additional control will be useful for some people.” He also announced that the “Rate limits are now 3,000 messages/week with GPT-5 Thinking, and then extra capacity on GPT-5 Thinking mini after that limit.”

In addition, the context window for GPT-5 Thinking is 196k tokens, which is good news for those using it for large coding projects. Altman iterated that they might increase the rate limits later down the line, depending on the usage patterns.

Am I the only one who feels like GPT-5 is colder and less empathetic?
byu/supream66 inChatGPT

ChatGPT Plus and Pro subscribers will also see a “Show additional models” toggle in the ChatGPT web settings, which will bring back models like OpenAI o3, GPT-4.1, and GPT-5 Thinking mini. On the other hand, GPT-4.5 will remain exclusive to Pro users since it uses more GPUs to run.

This slate of changes comes in response to several complaints raised by Reddit users who said that GPT-5 responses lack a conversational tone and feel somewhat cold, compared to the older GPT-4o model. With these updates, OpenAI aims to address these concerns to make interactions feel more natural and engaging for users.