Right-Sizing Intelligence: Glean Saves 81% on Token Costs and Is Preferred 78% of the Time Versus Claude Cowork
Glean, Wednesday, August 26th, 2026
Glean reports 81% lower token costs and 78% preference against Claude Cowork through right-sizing model selection.
Glean has published results from an evaluation comparing its approach against Claude Cowork. The company reports saving 81% on token costs while being preferred 78% of the time in the comparison.
The argument behind the numbers is right-sizing intelligence: routing each request to the appropriate model capability rather than sending everything to the most capable and most expensive option.
Glean's position is that most enterprise requests do not require frontier reasoning, and that matching model to task is where both cost and quality gains come from.