OpenAI CEO Sam Altman has confirmed that ChatGPT Plus users can continue using the GPT-4o model if they prefer it over the new GPT-5, launched yesterday. The decision follows user backlash after GPT-4o was replaced without warning. Many users had grown used to GPT-4’s tone, personality, and capabilities, which they felt GPT-5 failed to match. Complaints also centered on GPT-5’s shorter responses, reduced task performance, and quicker usage limit hits.
According to the numerous Reddit users, they voiced dissatisfaction with GPT-5. They claimed it lacked GPT-4o’s personality and struggled to complete tasks at the same standard. Some users also said they reached GPT-5’s usage limits faster than expected.
In response, Altman announced that Plus users could choose to stick with GPT-4o. OpenAI will monitor usage before deciding how long to support older models. Additionally, GPT-5 rate limits for Plus users will be doubled as the rollout continues. The GPT-5 deployment is taking longer than planned, and some users still cannot access it. Altman noted that GPT-5 should feel smarter over time. An autoswitcher bug yesterday made GPT-5 appear “way dumber” than intended.
Altman emphasized that user feedback will remain central to future updates. Starting with iOS 26, Siri’s ChatGPT integration will use GPT-5. Until then, Siri will continue relying on older ChatGPT models.