Skip to content

OpenAI

OpenAI pulls GPT-5.5 from Codex on October 14, API spared

Written by an AI model from the linked sources and checked against them by a second model. How we work

All about: Coding agents →
Image: gizmoindo.com

OpenAI is switching GPT-5.5 off in ChatGPT, ChatGPT Work and Codex on October 14, across every plan, the company said in a September 15 post on its ChatGPT account on X. The OpenAI API is not part of the retirement: anything authenticated with an API key keeps working as usual, Gizmochina says in its write-up.

Most ChatGPT users have nothing to do. Their conversations move to whatever model OpenAI sets as the new default, and OpenAI has not named that model. Codex users who sign in with a ChatGPT account do have work: the post points them to GPT-5.6 Sol or GPT-6 Astra. Switching means updating workspace defaults, saved model settings, managed configurations, custom agents, scheduled tasks and any script that calls GPT-5.5 by name.

GPT-5.4 class models already left Codex at the end of August 2026. GPT-5.5 itself launched on April 23, 2026, less than two months after GPT-5.4, which CNBC called the latest sign of the breakneck pace of development in the AI sector. The retirement post went out twelve days after Astra’s September 3 release. Gizmochina reads the move as part of OpenAI’s recent pattern of retiring older models once newer ones are ready, while OpenAI frames it as keeping things current.

For production setups, the split by authentication is what decides the work. Calls made with an API key keep GPT-5.5 for now. Developers who hardcoded the model name into apps or automation tools that run through a ChatGPT sign-in face the most edits, and the model string in those has to point to GPT-5.6 Sol or GPT-6 Astra before October 14.

No comments yet. Be the first.

Join the conversation

Sign in to leave a comment. It appears after moderation.