Lates News
OpenAI: GPT-5 performs exceptionally well in answering health-related questions. In the "HealthBench Hard Hallucinations" test measuring the accuracy of AI models' responses on health topics, GPT-5 has a hallucination rate of only 1.6%. This is significantly lower than the company's previous GPT-4o and o3 models, which had hallucination rates of 12.9% and 15.8% respectively.
Latest