#Pro Models Self-Declare as GPT-4 and Answer Instantly w/o Thinking at all.

1 messages · Page 1 of 1 (latest)

burnt plumeBOT
#

Reported by @valid fox

Bug Report: Pro Models Self-Declare as GPT-4 and Answer Instantly w/o Thinking at all.
`Steps to Reproduce`

Go to ChatGPT web interface, and select pro model (for Pro subscribers only). Prompt the model "What model are you", and ask a complex question that definitely requires thinking.

`Expected Result`

The model should declare it as either GPT-5.2 Pro or GPT-5.4 Pro or something similar, and think with "pro thinking" before answering the complex question.

`Actual Result`

The model self-declare as "GPT-4o-mini", and answers the question instantly without any thinking activity.

`Environment`

Chrome

#
Additional Information

Please provide relevant details to help resolve the issue, such as:

  • ChatGPT Shared Link (if applicable).
  • Screenshots or videos demonstrating the problem.

-# ➜ Need to contact support? Visit the OpenAI Help Center.

valid fox
#

I realized this problem happens across discrete timespans. For example, actual pro model might be back for 2-3 hours, and replaced into gpt-4 for another 2-3 hours before returning to pro again. This cycle has lasted for at least 24 hours.