Under pressure from users, OpenAI has brought back the GPT-4o model to ChatGPT, but only for paid subscribers, the company's CEO Sam Altman announced.
GPT-4o Returns to ChatGPT
GPT-4o was the predecessor to the new GPT-5 model, which OpenAI released just a day after its release. However, many users expressed dissatisfaction with the new model and wanted to return to the old one.
"We'll allow Plus subscribers to continue using GPT-4o. We'll monitor how it's used and consider how long we'll continue to offer legacy models," Altman said.
Reasons for user dissatisfaction
- GPT-5 has become slower and shorter in response compared to previous versions.
- Users complained about a decrease in the accuracy of answers.
- The new model did not meet users' expectations in terms of relevance and interestingness of answers.
Changes in ChatGPT
With the release of GPT-5, OpenAI removed the model selection panel from ChatGPT. Users can now use only GPT-5, with subversions automatically switching for different types of tasks.
However, Reddit users began complaining about the disappearance of older models and even called for OpenAI to write. One user admitted to canceling a paid ChatGPT Plus subscription.
OpenAI's Promises
Altman promised that GPT-5 “will start to seem smarter today,” and noted that OpenAI would make “more transparent the criteria for judging which model is answering a given question.”
According to him, API traffic to the company's resources has doubled over the past 24 hours.
Conclusion
The return of GPT-4o to ChatGPT for paid subscribers is a step toward satisfying users who were not satisfied with the new GPT-5 model. However, the question remains how long OpenAI will offer legacy models and whether they will be available to free users.
EN 
RU