OpenAI Reorganises Stack
- OpenAI retired several ChatGPT models from the consumer chat product while keeping API access and enterprise entitlements intact. - It is rolling ads into Free and Go plans, preserving ad-free Pro/Business/Edu tiers, and updating fallback models for cost-performance. - Microsoft is adding OpenAI's GPT-image-2 into Microsoft Foundry, showing image-generation being packaged for enterprise design and localisation workflows. ( )
OpenAI has stripped several older models out of ChatGPT, started testing ads on cheaper plans, and kept selling newer models through business channels. (help.openai.com) OpenAI said 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. The company said API access to those models was unchanged. (help.openai.com) For workplace customers, the cutoff came later: OpenAI said ChatGPT Business, Enterprise, and Edu customers kept GPT-4o inside Custom GPTs until April 3, 2026. OpenAI’s Help Center now says GPT-4o is fully retired across plans after that date. (help.openai.com) OpenAI also changed what happens when users hit limits. In ChatGPT release notes dated April 9, 2026, the company said GPT-5.3 Instant Mini replaced GPT-5 Instant Mini as the fallback model after users exhaust GPT-5.3 Instant usage. (help.openai.com) At the same time, OpenAI is testing advertising inside ChatGPT. Its ads policy says ads may appear for Free and Go users in the United States, Australia, New Zealand, and Canada, while Plus, Pro, Business, Enterprise, and Edu accounts remain ad-free. (help.openai.com) Go is OpenAI’s lower-cost paid tier, and the company says it is now available in all ChatGPT-supported countries. That gives OpenAI a three-part consumer ladder: ad-supported Free and Go plans, ad-free higher tiers, and separate API access for developers. (help.openai.com; help.openai.com) Microsoft is packaging the same model reshuffle for enterprise buyers. On April 21, 2026, Microsoft said OpenAI’s GPT-image-2 was generally available and rolling out in Microsoft Foundry, its platform for building and managing artificial intelligence applications. (techcommunity.microsoft.com) Microsoft said GPT-image-2 is aimed at teams making large batches of marketing and product images, with multilingual understanding, higher-resolution output, editing controls, and a routing layer that steers requests for production use. The company framed it around localized campaigns, product catalogs, and brand-design workflows. (techcommunity.microsoft.com) That fits Microsoft’s broader push to sell OpenAI models as enterprise infrastructure rather than consumer features. In March, Microsoft added GPT-5.4 and GPT-5.3 Chat to Foundry for production workloads and high-volume customer conversations. (techcommunity.microsoft.com; techcommunity.microsoft.com) The result is a cleaner split: ChatGPT gets fewer visible model choices and more pricing tiers, while OpenAI and Microsoft keep widening the menu for developers and corporate buyers. (help.openai.com; help.openai.com; techcommunity.microsoft.com)