New AI Model BDH-CQ Claims to Run 11 Times Cheaper Than a Leading OpenAI Model

A new artificial intelligence model has been designed to dramatically cut the cost of running AI requests using a novel approach to machine cognition, and its creators suggest nonverbal reasoning could be a step toward more human-like machine intelligence. Scientists at the AI company Pathway detailed the technical foundations of the model, called BDH-CQ, in a paper published Aug. 10 on the preprint server arXiv.

The paper reports that BDH-CQ costs up to 11 times less to run than a leading OpenAI model. Because the study has been posted as a preprint rather than published in a peer-reviewed journal, and the cost comparison comes from Pathway's own paper, the claim has not yet been independently verified.

BDH-CQ builds on an earlier model called "Dragon Hatchling," which the same scientists created in 2025 to accurately simulate how neurons in the brain connect and strengthen during learning.