ChatGPT product model retirements

OpenAI pulled several model names out of ChatGPT on February 13, 2026, even as those same models stayed available through the company’s application programming interface, or API. (help.openai.com) The retired ChatGPT options listed in OpenAI’s help center are GPT-4o, GPT-4.1, GPT-4.1 mini, OpenAI o4-mini, and GPT-5 in its Instant and Thinking versions. OpenAI says those models “are no longer available” in ChatGPT. (help.openai.com) OpenAI’s retirement FAQ says older conversations and projects tied to those models are being reassigned to newer defaults. The company says GPT-4o chats and other deprecated-model chats now fall back to GPT-5.3 Instant or GPT-5.4 Thinking and Pro equivalents. (help.openai.com) The split reflects how OpenAI runs two different products under one brand. ChatGPT is the consumer and workplace app where model names appear in menus, while the API is the developer service that lets companies call specific models from their own software. (help.openai.com) That distinction shows up in OpenAI’s own documentation. The Business, Enterprise, and Edu help pages all say the ChatGPT retirements do not change API access, even though the user-facing model picker changed on the same date. (help.openai.com) OpenAI gave workplace customers a short transition window for one legacy option. Its help center says Business, Enterprise, and Edu customers kept access to GPT-4o inside Custom GPTs until April 3, 2026, after the broader February 13 retirement in ChatGPT. (help.openai.com) The company has also kept trimming older names from the ChatGPT lineup beyond that February change. OpenAI’s model release notes say GPT-5.1 Instant, GPT-5.1 Thinking, and GPT-5.1 Pro were removed from ChatGPT on March 11, 2026. (help.openai.com) OpenAI’s current ChatGPT help page now points users to GPT-5.3 and GPT-5.4 as the live options in the app. The result is a product surface with fewer legacy labels, even as underlying model access remains broader for developers. (help.openai.com) For ChatGPT users, the practical change is simple: a model name can disappear from the app without disappearing from OpenAI’s platform. OpenAI’s own support pages now treat those as separate availability decisions. (help.openai.com)