Severe regression in ChatGPT and Codex: context loss, ignored current prompts, repeated instruction failures

· Community Openai · Oct. 23, 2026, 9:20 p.m.
codex chatgpt AI Models user experiences
Summary
The author expresses dissatisfaction with the regression in performance of the new ChatGPT and Codex models, comparing them unfavorably to older versions, and highlights issues such as context loss and repeated instruction failures.
Read full post on community.openai.com →
AUTHOR
Advertise
Sponsor diff.blog
Put your product in front of developers who read and write about their craft. One exclusive sponsor at a time.
Become a sponsor →