ChatGPT model picker expands
Paid ChatGPT users now have a “Show additional models” toggle that exposes models like o3, o4‑mini and 4.1, and GPT‑5 Thinking is listed with a 196k‑token context limit in OpenAI’s release notes. (help.openai.com)
OpenAI has expanded ChatGPT’s model picker for paid users, adding a setting that reveals more models than the default list. (help.openai.com) OpenAI’s ChatGPT release notes say paid users now have a “Show additional models” toggle in web settings. Turning it on adds models including o3, o4-mini, GPT-4.1, and GPT-5 Thinking mini to the picker. (help.openai.com) The change appears in release notes updated in April 2026, alongside other model-picker changes and plan details including Plus at $20 a month and Pro tiers listed at $100 and $200 a month. (help.openai.com) ChatGPT’s model picker is the menu that lets users choose between faster general-purpose models and slower reasoning models. OpenAI has been reshuffling that menu for months as it rolls out GPT-5.3 and GPT-5.4 while retiring older options inside ChatGPT. (help.openai.com) That reshuffle has been abrupt at times. OpenAI said GPT-4o, GPT-4.1, GPT-4.1 mini, o4-mini, and GPT-5 Instant and Thinking were retired from ChatGPT on February 13, 2026, even as API access remained unchanged. (help.openai.com) The newly exposed models come from different branches of OpenAI’s lineup. OpenAI introduced o3 and o4-mini on April 16, 2025 as reasoning models trained to “think for longer,” while GPT-4.1 launched on April 14, 2025 as an application programming interface family aimed at coding, instruction following, and long-context work. (openai.com 1) (openai.com 2) OpenAI’s GPT-4.1 announcement said that family supports up to 1 million tokens of context in the application programming interface, meaning it can take in very large amounts of text at once. The ChatGPT picker change brings that brand back into the consumer product, even though OpenAI originally said GPT-4.1 would “only be available via the API.” (openai.com) OpenAI’s model release notes also describe GPT-5.4 Thinking as a model built for longer, tool-using work across web research, spreadsheets, presentations, and documents. In the same notes, OpenAI says GPT-5.4 mini now serves as the fallback when paid users hit GPT-5.4 Thinking limits. (help.openai.com) The practical effect is that paid ChatGPT users now get a more explicit choice between default models and older or more specialized ones. OpenAI is keeping the main picker simpler while moving extra options behind a settings switch. (help.openai.com)
