The tests reveal that the chatppt-5 hallmine less than GPT-4O-and Grok is always the king to invent stuff


  • Chatgpt 5 scores of 1.4% on the hallucination classification
  • This places it before the Chatppt-4 which marks 1.8% and GPT-4O, which marks 1.49%
  • Grok 4 is much higher at 4.8% with Gemini-2.5 Pro is 2.6%

When Openai launched Chatgpt-5 Thursday last week if the major sales arguments that CEO Sam Altman underlined was that Chatgpt-5 was the most powerful, intelligent, fastest, reliable and robust version of Chatgpt that we have never been shipped “, and in the presentation, Openai staff also stressed that the Chatgpt-5” attenuated the hallucinations ยป.

When AI invents something, it is called a hallucination, and although hallucination rates decrease among all LLM, it is always surprisingly common, and one of the main reasons that we cannot trust AI to perform a task without human supervision.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top