author: "Matthew Zhao"

Glean's platform uses an automatic routing system to select the most appropriate and cost-effective AI model for different enterprise tasks. A benchmark of over 180 business tasks showed this approach significantly reduces expenses. By intelligently routing queries to smaller, more efficient models when possible, the system achieved an 81% reduction in token costs compared to exclusively using a large model like Claude Cowork. Furthermore, users preferred the output from Glean's routing system 78% of the time, highlighting that high-quality context can deliver superior results even with less powerful models.