OpenAI Launches GPT-4.5: A Leap in AI Tech

OpenAI has unveiled GPT-4.5, promising significant improvements in AI accuracy and context understanding. The model reduces errors and is already available to ChatGPT Pro users.


OpenAI Launches GPT-4.5: A Leap in AI Tech

OpenAI has presented GPT-4.5, the most advanced version of its artificial intelligence (AI) model to date. Highlighting significant improvements in precision, context understanding, and error reduction compared to its predecessor, GPT-4, the company shared the news through its social media, accompanied by an explanatory video about the model's new features and capabilities.

According to the company, one of the most notable improvements of GPT-4.5 is the significant reduction in "hallucinations," a term used to describe the generation of inaccurate information by the model. In internal tests, GPT-4.5 demonstrated a hallucination rate of 37%, a considerable improvement over the almost 60% recorded by its predecessor. Mia Glaese, a research scientist at OpenAI, mentioned in a video released by the company that "the model has increased its overall precision, providing more reliable answers in various areas of knowledge."

Additionally, GPT-4.5's improved ability to understand and generate text with greater coherence and contextualization in its responses is emphasized. OpenAI stresses that this model possesses a broader knowledge of the world and can respond with greater depth, although this has led to an increase in operational costs due to the higher demand for computational processing.

The initial availability of GPT-4.5 is limited to users with a ChatGPT Pro subscription of $200 per month, but OpenAI has announced that the model will be gradually implemented for a larger number of users starting next week.

The launch of GPT-4.5 takes place in a context of intense activity in the AI industry, with competitors like Anthropic and xAI presenting their own advanced models. Additionally, start-ups like DeepSeek in China have developed smaller and more accessible models, creating pressure on OpenAI to continue innovating in large-scale models.

In Nuevo León, the impact of GPT-4.5 is reflected in the educational sphere, where institutions like the Tecnológico de Monterrey have implemented specialized courses to optimize the use of ChatGPT. This provides students and professionals with tools to leverage this artificial intelligence in academic and work environments.