GPT-4o Leaves ChatGPT — Product Hygiene Locks Consumers onto 5.x
OpenAI (Jan 29) said GPT-4o, GPT-4.1, GPT-4.1 mini, and o4-mini leave ChatGPT on Feb 13—with previously announced GPT-5 Instant/Thinking retirements. ~0.1% of users still choosing GPT-4o daily; migration to GPT-5.2. API availability unchanged at announcement. Fewer serving stacks, clearer default UX.
Model retirement is product hygiene and cost control: fewer parallel serving stacks, clearer defaults, pressure on users pinned to older “personality” models. OpenAI’s proprietary close of the post-GPT-4 consumer chapter: ChatGPT locks onto the 5.x generation.
OpenAI on January 29 announced that GPT‑4o, GPT‑4.1, GPT‑4.1 mini, and OpenAI o4-mini will leave ChatGPT on February 13, 2026, joining the already planned retirement of GPT‑5 Instant and Thinking. Rationale: improvements on newer models; usage concentrated on GPT‑5.2; only ~0.1% of users still picking GPT‑4o each day (company). API model availability unchanged at announcement—developers continue under separate deprecation schedules if any. Roughly two weeks for power users still pinned to 4o-class pickers to migrate habits.
Claims vs checks
Retirement list, effective date, and 0.1% figure are OpenAI primary. “Vast majority on 5.2” is company usage telemetry—not a public dashboard. API “no change” is announcement-time only; watch separate deprecation notices.
Limits
- ChatGPT picker ≠ API catalog.
- Personality preference is real for a small cohort—0.1% can still be loud.
- Downstream apps hard-coded to 4o chat behavior need their own migration plans.
Sources
- OpenAI: “Retiring GPT‑4o, GPT‑4.1, GPT‑4.1 mini, and OpenAI o4-mini in ChatGPT” (January 29, 2026). Primary.