# Beyond Round Robin: Building a Token-Aware Load Balancer for LLMs

In my previous experiment, I was trying to find the best model for a given task. The approach was to...