OpenAI promised developers six months’ notice. GPT-4o users got three.
OpenAI is retiring four GPT-4 models from its API on January 14, 2027. The deprecation notice the company sent developers The post OpenAI promised developers six months’ notice. GPT-4o users got three. appeared first on The New Stack .
OpenAI announced the retirement of four GPT-4 models from its API on January 14, 2027. The company informed developers about the deprecation of GPT-4o, GPT-4o mini, GPT-4.1, and GPT-4.1 mini, along with their snapshots and fine-tuned models, giving them only three months to migrate production traffic. The recommended replacement, GPT-6 Luna, handles reasoning and tool calls differently enough that changing the model ID may be the easiest part of the migration process.
GPT-4o debuted on May 13, 2024, and gained a loyal following, leading OpenAI to bring it back for paying ChatGPT subscribers just a day after its replacement, GPT-5, was introduced in August 2025. However, GPT-4o and GPT-4.1 were sunsetted from ChatGPT by February 13, 2026. While GPT-4.1 was initially an API-only model, it later became available to developers with enhanced coding abilities and a larger context window for token processing.
Other GPT-4 models, including GPT-4.1 nano, GPT-4o snapshots, and older generations, are also set to retire on various dates leading up to January 14, 2027. The migration guide suggests using Luna's default medium reasoning effort, but function calling through Chat Completions requires using the Responses API. Fine-tuning jobs on Luna will end on January 6, 2027, leaving a brief window to replace these models.
OpenAI's deprecation policy typically provides six months' notice for generally available models, but in this case, GPT-4o and GPT-4.1 received only three months' warning, raising questions about the explanation for this shortened notice period.
Written by urgent.news from The New Stack's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
