OpenAI has unveiled its latest AI model, GPT-4.5, code-named Orion, marking a significant advancement in conversational AI. Unlike its predecessors, GPT-4.5 leverages unsupervised learning to enhance its world knowledge and emotional intelligence, offering users a more natural and intuitive interaction experience.
What’s New in GPT-4.5?
GPT-4.5 focuses on improving emotional intelligence and understanding human intent, making it better suited for both professional tasks and creative endeavors. Testers have praised the model for its reduced hallucinations and more accurate responses, making it a reliable tool for everyday queries.
Performance and Benchmarks
While GPT-4.5 doesn’t show significant improvements in math or science, it outperforms previous models like o3-mini and o1 on OpenAI’s new SWE-Lancer benchmark, designed for freelance coding tasks. This makes it a strong contender for developers and tech professionals.
Access and Pricing
Currently, GPT-4.5 is available exclusively to Pro users and developers on paid plans, with Plus and Team users gaining access next week. However, the API pricing has raised eyebrows, costing $75 per million input tokens and $150 per million output tokens — a stark contrast to GPT-4o’s $2.50/$10 pricing.
Is GPT-4.5 Worth It?
While GPT-4.5 delivers a more engaging conversational experience, its high costs and incremental improvements may leave some users underwhelmed. This release may represent a plateau in non-reasoning model development, signaling a shift in OpenAI’s focus.
For now, GPT-4.5 is a personality upgrade rather than a groundbreaking leap, making it best suited for those prioritizing emotional intelligence and conversational fluidity over technical advancements.