OpenAI has announced the release of its new flagship model, GPT-5, which will replace all existing models in the ChatGPT platform, including GPT-4o and others. This decision was shared by OpenAI co-founder and CEO Sam Altman, who stated that the rollout of GPT-5 was “bumpier” than anticipated. Selected users will regain access to older models for a period of 60 days, but the transition to GPT-5 will ultimately eliminate the option for users to select previous models like GPT-4o, o3, o4-mini, or o4-mini-high.
The rollout of GPT-5 will be available to ChatGPT Plus, Free, Pro, and Team users. However, organizations subscribed to the Enterprise and Edu tiers will still have access to legacy models for a limited time. Many users, accustomed to the older models for daily tasks, have expressed concern over this sudden change. The transition might impact workflows that were built around the previous models’ responses and capabilities.
This has not been the first instance where OpenAI has sunsetted models, but it marks the first comprehensive action of this nature within the ChatGPT platform. Prior to this, Altman had acknowledged the confusion surrounding the model selection process and hinted at a desire to streamline user experience. Some anticipate that forcing users to adopt the latest model may lead to greater engagement with newer capabilities.
OpenAI has clarified that while older models will be deprecated on the ChatGPT platform, they do not plan to discontinue older models on the API side. This reassures enterprises that their applications will remain unaffected by the sunsetting of legacy models on the ChatGPT interface. The company’s approach seems to balance the need for innovation while providing a safety net for enterprise clients relying on existing models for application development.
Source: https://venturebeat.com/ai/chatgpt-users-dismayed-as-openai-pulls-popular-models-gpt-4o-o3-and-more-enterprise-api-remains-for-now/