Following the unpopular launch of GPT-5, OpenAI has begun restoring optional access to previous artificial intelligence models in a swift retreat after thousands of users complained about losing access to their preferred models.
Article contents:
- Introduction: A swift retreat from OpenAI
- User revolt: “GPT-5 is terrible”
- Addressing user concerns: More control and options
- “Personality” concerns: From conversational to abrupt
- A plagued launch
- The path forward: Balancing innovation and choice
- Conclusion: The user’s voice is heard
Introduction: A swift retreat from OpenAI
OpenAI CEO Sam Altman announced on Tuesday that GPT-4o has returned to ChatGPT following a fierce user backlash over its removal during the launch of GPT-5 last week. The AI model now appears in the model picker for all paid ChatGPT users by default, marking a rapid reversal after thousands of users complained about losing access to their favorite models.
User revolt: “GPT-5 is terrible”
The return of GPT-4o comes after what Altman described as OpenAI underestimating “how much some things people loved about GPT-4o mattered to them.” In an effort to streamline its offerings, OpenAI initially removed all previous AI models from ChatGPT when launching GPT-5 on August 7, forcing users to adopt the new model without prior warning. This move sparked one of the most vocal user revolts in ChatGPT history, with a Reddit thread titled “GPT-5 is terrible” gathering over 2,000 comments within days.
Addressing user concerns: More control and options
Alongside bringing back GPT-4o, OpenAI has implemented several other changes to address user concerns. Rate limits for the reasoning mode in GPT-5 were increased from 200 to 3,000 messages per week. The company also added new routing options—”Auto,” “Fast,” and “Reasoning”—giving users more control over which GPT-5 variant handles their queries. Pro users will also gain access to other additional models.
“Personality” concerns: From conversational to abrupt
In addition to model availability, OpenAI acknowledged widespread complaints regarding the output style of GPT-5. Users described the new model as “abrupt and sharp” compared to the more conversational tone of GPT-4o. Some users with emotional attachments to the older model expressed grief over losing what they considered their “only friend.”
Altman wrote in his announcement: “We are working on a GPT-5 personality update that should feel warmer than the current one but not as annoying (to most users) as GPT-4o.” He added that OpenAI recognizes the need for “more per-user customization of the model’s personality” moving forward.
A plagued launch
The launch of GPT-5 was fraught with multiple issues aside from the model removal. An auto-routing system intended to determine appropriate model variants failed on launch day, repeatedly falling back to less capable versions. OpenAI also faced criticism for including misleading performance charts in the launch presentation, which Altman later described as a “bad chart error.”
The path forward: Balancing innovation and choice
While GPT-4o is back for now, OpenAI continues to fine-tune GPT-5. The company indicated that rate limits may need further adjustments “depending on usage” and that personality updates are still under development. For now, paid ChatGPT users can select their preferred model, providing a compromise between OpenAI’s push toward newer AI models and user demands for choice.
Conclusion: The user’s voice is heard
OpenAI’s rapid response to the user backlash demonstrates how vital the user community is to the company. In a competitive AI market, companies can no longer force changes on users without consequences. This incident serves as a strong reminder that user experience and preferences still play a critical role in the success of any technology product, even for industry leaders like OpenAI.
