OpenAI, the company behind ChatGPT, has announced the retirement of several legacy AI models from ChatGPT, effective February 13, 2026. This retirement affects multiple models, including GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini. This means that users will now be flung fully into the GPT 5 series of models.
The decision to retire these older models coincides with the previously scheduled retirement of GPT-5 (Instant and Thinking) versions on the same date. By consolidating its available models, the company wants to focus on the development and maintenance of its GPT 5.2 models, which most people are using.
While ChatGPT users will lose access to these models, there are no changes or retirements to these models within the API platforms, so third-party developers can continue to use them in their apps.
GPT-4o was a big deal when it launched in May 2024 for free users. It bridged the intelligence gap between GPT 3.5 and the smarter GPT-4 which was for paid users only for a substantial amount of time. Nowadays, OpenAI says it is used by only 0.1% of daily users, however, it was kept around for Plus and Pro users who valued its conversational warmth.
GPT-5 launched in the middle of last year and it soon drew criticism from users who noticed a personality shift. OpenAI improved this with GPT-5.1 by making the model warmer and more friendly. Now that it has made this change, it no longer needs to keep GPT-4o around for its personality.
OpenAI is expecting backlash from this change but plans to release further updates shortly that address creative personality and refinement of the model’s refusal logic. These patches will help to fill the gap left by 4o and the other models.