OpenAI has released GPT-4.5, its largest and best chat model yet, as a research preview.

What is GPT-4.5?

GPT-4.5 is not a product for general use, but a research preview to better understand its strengths and limitations. It is OpenAI's "last non-chain-of-thought model" — that is, it contains no explicit chain-of-thought (CoT) mechanisms in its training, according to reports from tech journalists.

Core Contributors

A roster of notable researchers stand behind the model:

  • Aiden Low, Alec Radford, Alex Carney, Alex Nichol
  • Alexis Conneau, Ananya Kumar, Ben Wang
  • Charlotte Cole, Elizabeth Yang, Gabriel Goh
  • Hadi Salman, Haitang Hu, Heewoo Jun
  • Ian Sohl, Ishaan Gulrajani, Jacob Coxon
  • James Betker, Jamie Kiros, Jessica Landon
  • Kyle Luther, Lia Guy, Lukas Kondraciuk
  • Lyric Doshi, Mikhail Pavlov, Qiming Yuan
  • Reimar Leike, Rowan Zellers
  • Sean Metzger, Shengjia Zhao, Spencer Papay
  • Tao Wang

Plus many more contributors including Adam Lerer, Adrien Ecoffet, Aidan McLaughlin, Alexander Prokofiev and many others.

Pricing

According to an Actuia report from October 2026:

  • GPT-4.5 research preview: $0.50 – $50 per output
  • Available through OpenAI's API

Who Should Read This?

  • AI developers wanting to test model limits
  • Researchers studying non-CoT models
  • Enterprise customers seeking alternatives to chain-of-thought approaches

Conclusion

GPT-4.5 represents a generational leap for chat models without explicit CoT training. As a research preview, it opens new possibilities for testing model capability in unexpected scenarios.

Sources: