As reported by CNBC, on January 30, 2026 (Beijing Time), OpenAI made an announcement that it would start discontinuing several models within its ChatGPT chatbot from February onwards. Among these models is the once highly popular GPT-4o. Launched in May 2024, GPT-4o swiftly won the hearts of a portion of paid users due to its amiable conversational tone. Nevertheless, following the introduction of GPT-5 in August 2025, OpenAI temporarily restricted access to GPT-4o. This move triggered a wave of discontent among users, although access was promptly reinstated. OpenAI explained that at present, a mere 0.1% of users opt for GPT-4o, with the vast majority having shifted to the more sophisticated GPT-5.2. Consequently, the company has made the decision to discontinue older models such as GPT-4o in order to focus its resources on enhancing mainstream models. It's important to note that this discontinuation is applicable solely to the ChatGPT client, leaving API services untouched. Developers will still have the option to continue utilizing the aforementioned retired models through the API.