Model routing
Model routing is the practice of sending each AI request to a model chosen for that request, instead of sending everything to one fixed model. The choice is made by rules, such as task type, team or data class, or automatically by a classifier.
Why it matters for companies
Prices per token differ widely between models. Routing routine work to cheaper models is one of the largest levers on an AI bill.
Routing happens at two levels. The same model is often served by several providers with different prices, speeds, locations and contracts, so the provider choice matters as much as the model choice.
Example
A support team's ticket triage goes to a fast, low-cost model, draft replies to customers go to a frontier model, and anything tagged as personal data goes only to models on European endpoints.
How Standard Compute handles it
A single API gives you frontier and open-source models, and it routes each step to the cheapest model that can do it and the cheapest provider fast enough to serve it.
Decide which models each key can use, so every app and every team member gets exactly the models you approve.