OpenAI has taken a new step in the evolution of artificial intelligence with the launch of GPT-5.4, just a day after introducing GPT-5.3 Instant. This new model, announced as the most advanced and efficient for professional environments, promises to transform the way we interact with digital tools, from spreadsheets to financial analysis. GPT-5.4 stands out as the first OpenAI model with native computing capabilities, allowing it to execute multiple tasks simultaneously across different applications. This functionality, known in other systems as "Computer," is now more seamlessly integrated into ChatGPT, enabling the AI to operate as if it were controlling a real computer and executing the necessary steps to fulfill user requests. One of the most notable improvements is its ability to navigate desktop environments, recording keyboard and mouse commands with greater accuracy than previous versions. Additionally, GPT-5.4 becomes the default model in ChatGPT's "Thinking" mode, showing users the steps it will take before executing a command, facilitating real-time intervention and adjustment. OpenAI claims that GPT-5.4 is particularly effective for internet research, especially for very specific queries. According to the company, this model provides faster, more relevant, and higher-quality responses, with an 18% lower likelihood of making errors compared to GPT-5.2. Another significant innovation is the expansion of the context window, which now reaches one million tokens, allowing for more efficient processing and analysis of large volumes of information while using fewer resources than its predecessors. The accelerated release of new models by OpenAI reflects the intense competition in the sector, with rivals like Google and Anthropic driving innovation. Each OpenAI model is optimized for specific tasks, making it crucial to choose the right one to maximize the benefits of a ChatGPT subscription. Source: elespanol.com