On October 14, the GPT-5.5 model was formally phased out from ChatGPT, ChatGPT Work, and Codex, all of which are accessible via ChatGPT accounts. However, it's important to note that related API services have not been impacted by this change. In the meantime, certain versions of several models, including GPT-4o, are set to be removed on January 14, 2027. This timeline grants developers around three months to transition to newer models. OpenAI's latest offerings, such as GPT-6.1, showcase enhancements in response speed and factual precision, coupled with more competitive pricing for API calls.

Currently, ChatGPT boasts a staggering 1.2 billion weekly active users. This immense user base necessitates OpenAI to strike a delicate balance between allocating computational resources for model training and providing user inference services. At the same time, OpenAI is locked in a fierce rivalry with Anthropic, competing on model capabilities, speed, and cost. Both companies have been actively iterating their models in recent times to stay ahead.

However, this competitive landscape has not been without its challenges. Multiple service disruptions have underscored the paramount importance of infrastructure and operational stability. Following the retirement of GPT-5.5, OpenAI has provided official alternative model solutions. Users are strongly encouraged to update their old task model settings and select suitable models and inference intensities based on their actual needs, thereby boosting efficiency.