OpenAI Reorganizes ChatGPT Models

OpenAI removed several standalone model options from ChatGPT—GPT-4o, GPT-4.1, GPT-4.1 mini, o4-mini and GPT-5 variants—while saying API access remains unchanged. Business, Enterprise and Education customers will still have GPT-4o inside Custom GPTs, and custom actions for GPTs are rolling out on web for Plus, Pro and Team with Enterprise and Edu coming later. (help.openai.com) (help.openai.com)

OpenAI has collapsed ChatGPT’s model picker again, retiring several named options and steering users toward a smaller set built around GPT-5.3 Instant and GPT-5.4 Thinking. (help.openai.com) OpenAI’s help center says GPT-4o, GPT-4.1, GPT-4.1 mini, OpenAI o4-mini, and GPT-5 Instant and Thinking were retired from ChatGPT on February 13, 2026. A second wave followed on March 11, when GPT-5.1 Instant, GPT-5.1 Thinking, and GPT-5.1 Pro were also removed from ChatGPT. (help.openai.com) The company says those models are gone from ChatGPT, not from the application programming interface that developers use to build software. OpenAI’s retirement notice says the retired models “will continue to be available through the OpenAI API,” and says the company will give advance notice before future application programming interface retirements. (help.openai.com) Inside ChatGPT itself, OpenAI is replacing a list of separate names with a routing system that picks between models for you. Its GPT-5.3 and GPT-5.4 article says GPT-5.3 is the default for logged-in users, and that selecting “Instant” can automatically switch some harder requests over to GPT-5.4 Thinking. (help.openai.com) That changes how people use ChatGPT even if they never touch the application programming interface. Instead of choosing among GPT-4o, GPT-4.1, or o4-mini by name, users now see simpler labels such as Instant, Thinking, and Pro while OpenAI decides more of the underlying model behavior. (help.openai.com) OpenAI says old chats and projects are not disappearing with the retired models. Its retirement notice says conversations and projects tied to older models are being defaulted to GPT-5.3 Instant or GPT-5.4 Thinking and Pro equivalents, and GPTs using retired models are moved to the closest replacement. (help.openai.com) The exception lasted longest for workplace customers building custom bots. OpenAI’s model-limits pages said ChatGPT Business, Enterprise, and Education customers would keep GPT-4o inside Custom GPTs until April 3, 2026, even after the broader February 13 retirement inside ChatGPT. (help.openai.com) That grace period has now ended. OpenAI’s GPT-5.3 and GPT-5.4 article says that after April 3, GPT-4o is “fully retired across all plans,” closing the last named holdout from the older ChatGPT lineup. (help.openai.com) OpenAI is still expanding what custom bots can do, even as it trims the list of model names. Its GPT editor documentation says builders can create and edit GPTs only on the web, and its actions documentation says GPTs can connect to outside application programming interfaces through custom actions, although pro-series models are not supported for those actions. (help.openai.com 1) (help.openai.com 2) The result is a cleaner storefront and a messier translation layer underneath it. OpenAI is asking ChatGPT users to think less about model names, while developers and workplace admins still have to track which models remain available, where they run, and when the next retirement notice lands. (help.openai.com 1) (help.openai.com 2)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.