OpenAI pruning model lineup
OpenAI is retiring several older model variants from ChatGPT while promoting newer GPT‑5.3 and GPT‑5.4 options, though API access for enterprise customers remains separately managed. The change is presented as product consolidation and affects which model labels appear to end users versus what Business, Enterprise and Edu customers can access. (help.openai.com)
OpenAI has removed several older model choices from ChatGPT and is steering users toward GPT-5.3 as the default and GPT-5.4 for heavier reasoning tasks. (help.openai.com) OpenAI’s help center says GPT-5.3 now powers the main logged-in ChatGPT experience, while GPT-5.3 Instant is rolling out as the faster everyday option and GPT-5.4 Thinking is positioned for harder work. The same article says GPT-4o was set to be fully retired across all plans after April 3, 2026. (help.openai.com) Separate help pages for ChatGPT Business, Enterprise, and Edu say 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. Those pages also say application programming interface access remains unchanged. (help.openai.com, help.openai.com) That split is the key to the change: the model names a person sees inside ChatGPT are no longer the same thing as the models a company can still call through OpenAI’s application programming interface. OpenAI’s retirement notice says GPT-5.1 models left ChatGPT on March 11, 2026 but stayed available through the application programming interface. (help.openai.com) OpenAI is also thinning the visible picker, not just swapping engines. Its model release notes say GPT-5.4 mini can be used as a fallback when GPT-5.4 Thinking hits rate limits, but GPT-5.4 mini will not appear as a selectable model, and GPT-5 Thinking mini is being retired as a selectable option within 30 days. (help.openai.com, help.openai.com) The company has been making the same argument across its product pages: fewer labels, more automatic routing. OpenAI’s GPT-5.3 help article describes the default experience as “a single auto-switching system,” and its March 3 product post says the update targets the parts of ChatGPT people notice most in daily use, including tone, relevance, and conversational flow. (help.openai.com, openai.com) The company is still adding new variants behind the scenes. ChatGPT release notes published April 9, 2026 say GPT-5.3 Instant Mini replaced GPT-5 Instant Mini as the fallback after users hit rate limits, and that fallback does not appear in the model picker. (help.openai.com) For paying organizations, OpenAI has kept a narrower bridge to legacy behavior while the front-end lineup changes. The retirement notice says Business, Enterprise, and Edu customers kept GPT-4o inside Custom GPTs until April 3, 2026, even as the broader ChatGPT lineup was being cut back. (help.openai.com) The result is a simpler ChatGPT menu for most users and a more separate back-end model policy for developers and large customers. OpenAI’s own documentation now treats those as different layers of the product, with retirements in ChatGPT happening on one schedule and application programming interface access on another. (help.openai.com, help.openai.com)
