OpenAI Unveils GPT-4.5 ‘Orion’: A Smarter, More Thoughtful AI Model with Enhanced Efficiency and Accuracy

OpenAI has unveiled GPT-4.5, internally known as 'Orion,' its most advanced AI language model to date.  GPT-4.5 is designed to improve writing, problem-solving, and programming skills by providing a more natural conversational style and greater global knowledge.  OpenAI's CEO, Sam Altman, hailed the model as "thoughtful," claiming that it delivers truly intelligent suggestions, making encounters feel more human.  

 From a technical aspect, GPT-4.5 acts as a link between traditional AI and cutting-edge reasoning models.  It is over ten times more computationally efficient than GPT-4, resulting in faster responses and higher accuracy.  OpenAI finds a significant reduction in misinformation, with delusion rates falling from 60% in GPT-4o to 37% in GPT-4.5.  The model's increased emotional intelligence (EQ) and creativity make it an ideal tool for writing and creative jobs.  

 Despite its achievements, GPT-4.5 lacks multimodal features like Voice Mode, video, and screen sharing.  Furthermore, the higher processing demands make it more expensive to run.  Currently, GPT-4.5 is available as a research preview via OpenAI APIs.  Developers with paid API access will be able to use it beginning February 28, with ChatGPT Pro customers ($200/month) receiving priority access, followed by Plus and Team subscribers following week.  

 As OpenAI continues to innovate, their roadmap indicates that GPT-5 will incorporate more advanced reasoning models such as o3, thereby strengthening its dominance in the AI landscape.  Meanwhile, the company's finances remain solid, with its most recent funding round valuing it at $157 billion.

This article is based on information from The Tech Portal