OpenAI plans to retire several models from ChatGPT on February 13, including GPT‑4o, GPT‑4.1, and o4-mini, saying only 0.1% of users still choose GPT-4o
OpenAI announced it will retire several models from its ChatGPT chatbot next month, including its GPT‑4o model that is beloved by some users.
Context & Ripple Effects
OpenAI has already used ChatGPT retirements to consolidate its consumer model lineup: it planned to replace GPT-4 in ChatGPT with GPT-4o while retaining GPT-4 API access, and later signaled that selectable older models would be temporary. This announcement extends that earlier GPT-4 replacement strategy to several more models.
The decision matters because model choice is both a product-control mechanism and a user-facing feature. Even a small cohort of users selecting GPT-4o may have workflows or preferences tied to its behavior, making the transition a test of how much backward choice ChatGPT will preserve.
First-order effects
- ChatGPT users who deliberately select GPT-4o, GPT-4.1, or o4-mini must move to the models OpenAI keeps available after February 13; the announcement does not say those models’ API access will change.
- OpenAI reduces the number of consumer-facing models it must present, maintain, and support, following its prior warning that older-model access would be phased out.
Second-order effects
- Users with prompts, custom instructions, or evaluation routines tuned to the retired models will need to retest them on the remaining options, while teams that need stable model behavior may place greater weight on API-based controls or alternative providers.
- The retirement makes model selection less of a differentiator inside ChatGPT and increases the importance of the performance, reliability, and migration experience of the models that replace the removed options.
Third-order effects
- If recurring retirements become the norm, consumer AI access will increasingly resemble a managed service rather than a durable software version: providers set the upgrade cadence and users absorb behavior changes.
- That raises the strategic value of model-access governance and operational assurance, especially for organizations whose workflows depend on reproducible outputs rather than simply access to a capable chatbot.
The trend: Consumer AI platforms are consolidating model catalogs and treating model availability as a continuously managed product decision rather than a permanent user entitlement.