1 paper · 1 filter
Ameya Anjarlekar, Rasoul Etesami, R Srikant
We study personalized LLM routing, where a router selects among heterogeneous models with different costs and capabilities. The router must minimize cumulative inference cost while…