OpenAI sunsets popular ChatGPT models, users dismayed by GPT-5 transition
OpenAI has announced a significant shift for its ChatGPT platform: the newly released flagship model, GPT-5, will now exclusively power the conversational AI. This means the company will discontinue all existing models within ChatGPT, including popular choices like GPT-4o, o3, o4-mini, and o4-mini-high, with users no longer able to select a specific model for their tasks.
The transition is already underway for ChatGPT Plus, Free, Pro, and Team subscribers, with only users on the Enterprise and Edu tiers retaining access to these “legacy” models for a limited 60-day period. This decision caught many ChatGPT users by surprise, particularly those who had integrated specific models into their daily workflows and grown accustomed to their unique response patterns and interaction styles. For some, the adjustment is expected to be challenging, as their established routines were built around the nuances of their preferred AI.
The sentiment among many users ranges from disappointment to a sense of genuine loss. One user, Harry Horsperg, expressed sadness over the apparent phasing out of GPT-4.5, a model he found superior for textual and conceptual analysis compared to other GPT-4x series iterations. Similarly, saikalyan lamented the loss of O3, a model they relied on for a significant portion of their work. The distress was particularly acute over the discontinuation of GPT-4o, which served as ChatGPT’s default model. Many users either preferred it or simply never saw the need to switch, finding it perfectly adequate for their needs. Some even reported developing a “connection” to their chosen model, finding OpenAI’s demonstration during the GPT-5 announcement – where GPT-4o was asked to write its own eulogy – to be in poor taste. As one user, Meadowbrook, commented, the handling of the transition felt “graceless.”
This consolidation of models aligns with past statements from OpenAI CEO Sam Altman, who in February acknowledged that the platform’s model selection tool had become overly complicated. Altman had previously vowed to simplify the user experience, and this wholesale replacement of models by GPT-5 now appears to be the definitive solution to that complexity. Rumors had even circulated last month about OpenAI potentially introducing an automatic model router designed to select the most appropriate model for users based on their workload. While OpenAI has retired models before, this marks the first time all existing models on the chat platform will be entirely removed and replaced.
Yet, not everyone views this change negatively. A significant portion of the AI community sees the deprecation of older models and the automatic upgrade to GPT-5 as a strategic move to propel ChatGPT’s vast user base into the future. With an estimated 400 million users, many of whom never ventured beyond the default or older models, this forced transition ensures that everyone will now be utilizing OpenAI’s most advanced and capable AI. Indeed, critics of the “AI is dumb” sentiment often point out that such opinions frequently stem from users who stick with outdated or less powerful models, never realizing the full potential of newer iterations. By removing older options, OpenAI effectively compels users to experience the latest advancements, potentially unlocking significant capability jumps that many might otherwise have missed.
For enterprises, the situation is notably different. While individual or team-level users on the ChatGPT Enterprise tier will experience the 60-day grace period for legacy models, businesses that have built applications or agents using OpenAI’s API (Application Programming Interface) can largely rest easy. OpenAI has explicitly stated it currently has no plans to deprecate older models on the API side. The company has committed to providing advanced notice to developers should any such changes be considered. This distinction is crucial for enterprises, many of which regularly evaluate and even switch large language models to optimize performance and manage costs.
The decision to unify ChatGPT under GPT-5 marks a pivotal moment for OpenAI, balancing the immediate disquiet of a vocal user base with a strategic vision for a more streamlined, universally advanced AI experience. It is a bold step that, while causing some initial disruption, aims to ensure all ChatGPT users are consistently leveraging the cutting edge of conversational AI.