Investing.com -- OpenAI, a leading AI research lab, has unveiled GPT-4.5, its largest and most knowledgeable model to date. This model builds on the capabilities of its predecessor, GPT-4o, and is designed to be more versatile. It has been trained using a combination of new supervision techniques, traditional methods like supervised fine-tuning (SFT), and reinforcement learning from human feedback (RLHF), similar to those used for GPT-4o.

OpenAI conducted thorough safety evaluations before deploying GPT-4.5 and found no significant increase in safety risk compared to existing models. Early testing indicates that interaction with GPT-4.5 feels more natural due to its broader knowledge base, stronger alignment with user intent, and improved emotional intelligence. These features make it well-suited for tasks like writing, programming, and solving practical problems, with fewer instances of hallucinations.

OpenAI is currently sharing GPT-4.5 as a research preview to better understand its strengths and limitations. The company is still exploring its capabilities and is eager to see how people use it in ways that may not have been anticipated.

OpenAI CEO Sam Altman stated on X that GPT-4.5 is ready and feels like talking to a thoughtful person. Altman expressed his astonishment at receiving genuinely good advice from the AI. However, he also noted that GPT-4.5 is a large, expensive model, and due to recent growth and a shortage of GPUs, its launch to plus and pro tiers had to be staggered. The company plans to add tens of thousands of GPUs next week and roll it out to the plus tier then. Altman also indicated that hundreds of thousands more GPUs will be coming soon.

Altman further clarified that GPT-4.5 is not a reasoning model and won’t outperform in benchmarks. He described it as a different kind of intelligence and expressed excitement for people to try it.

This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.