OpenAI reshuffles product access
OpenAI has retired several model variants from the ChatGPT consumer line but kept enterprise continuity by preserving GPT‑4o access for Business, Enterprise and Edu customers in custom GPTs and projects. At the same time OpenAI introduced a $100 Pro plan for Codex and expanded Codex availability, while Anthropic is moving certain third‑party integrations toward pay‑as‑you‑go billing — signalling a push toward metered pricing for heavy or developer‑focused AI use. (help.openai.com) (economictimes.indiatimes.com)
OpenAI just made the consumer version of ChatGPT simpler by taking several model names off the menu, while leaving a separate escape hatch for companies that had built workflows around an older model. In OpenAI’s help center, GPT-5.3 is now the default for logged-in users, and GPT-5.4 Thinking is positioned as the heavier reasoning option. (help.openai.com) The cleanup was bigger than a normal product rename. OpenAI 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, even though application programming interface access stayed unchanged. (help.openai.com) That split matters because ChatGPT and the application programming interface are now being treated like two different stores selling the same brand. Casual users get fewer knobs to turn inside ChatGPT, while developers can still call the retired models through the application programming interface. (help.openai.com) OpenAI did not cut off every business customer at once. Its workspace documentation says ChatGPT Business, Enterprise, and Edu customers retained GPT-4o inside custom GPTs until April 3, 2026, which gave companies a bridge for internal tools and classroom setups that were already built around that model. (help.openai.com) That tells you what OpenAI is protecting. If a student or office worker opens ChatGPT, OpenAI wants one main default model; if a company has a custom bot connected to documents, projects, or a teaching workflow, OpenAI is giving that setup a slower off-ramp. (help.openai.com) At the same time, OpenAI is carving coding into its own pricing lane. Multiple reports on April 10, 2026 said OpenAI introduced a $100-a-month ChatGPT Pro tier aimed at heavier Codex use, with 5 times the Codex usage of Plus and a temporary 10 times boost through May 31. (venturebeat.com) (cryptobriefing.com) OpenAI’s business page shows where that is heading. It now pitches agents like deep research and Codex alongside “unlimited access to GPT-5” for business users, plus the option to add credits as needed, which is software-company language for metering the expensive parts separately from the basic seat license. (openai.com) Anthropic is moving in a similar direction from the other side. Recent reporting says Claude subscribers can still use Claude models with third-party agents like OpenClaw, but that usage is being pushed toward pay-as-you-go or application programming interface billing instead of being covered by a flat subscription. (venturebeat.com) (thenextweb.com) So the new shape of the market is getting easier to see. The chatbot you open in a browser is being simplified into a cleaner consumer product, while the code-writing agent, the custom workplace bot, and the third-party integration are being priced more like cloud computing, where the heavier you use it, the more directly you pay. (help.openai.com) (openai.com) (venturebeat.com)
