Subquadratic
SubQ builds radically efficient AI models to make frontier intelligence affordable and accessible for workloads that demand massive context.
Publisher details
✕
Subquadratic (SubQ) is a frontier AI research company focused on making AI scale more efficiently through fundamental advances in model architecture. Rather than pursuing incremental improvements to conventional transformers, the company develops new approaches to reduce the computational costs of processing large amounts of information. Its team brings together researchers and engineers from organizations including Meta, Google, MIT, Oxford, and Cambridge, with the goal of translating architectural breakthroughs into practical AI products for enterprise workloads.