Skip to content

Model Serving & Economics

Model router

A system that sends each request to the most suitable model based on cost, quality, or task.

Example

A router sends simple questions to a cheap model and complex ones to a stronger model.

Why people use it

Teams use “Model router” when they need to balance model quality, reliability, speed, and cost.

What you'll hear

“How does Model router affect latency, quality, or cost at scale?”

Related terms