Glossary
What is TurboQuant? Definition and meaning
A quantization method from Google Research, published at ICLR 2026, that combines PolarQuant with a QJL residual step to compress vectors with little error.
Definition
TurboQuant: A quantization method from Google Research, published at ICLR 2026, that combines PolarQuant with a QJL residual step to compress vectors with little error.
In short: Google's low-bit compression of vectors and caches.
When it is used
Post-training quantization of models and caches for efficient deployment.
Learn more
The full guide to TurboQuant explains how it works, when to use it and how it compares with the alternatives, with examples. The glossary lists every term on the site from A to Z.