Model routing
Model routing is choosing which AI model handles a request based on factors like cost, speed, and task difficulty.
Model routing is the practice of sending each request to the AI model best suited for it instead of using one model for everything. It matters because you may get faster answers for simple tasks and stronger answers for harder ones, often at a lower cost. For example, a tool might route a short question like “What’s the weather?” to a small, quick model, but send a complicated request like “Summarize this contract and list risks” to a more capable model. Routing can happen automatically or be set by the tool’s maker.