Serving
Endpoint
What is endpoint?
An endpoint is a network address that serves a specific model, with its own scaling configuration and its own cost. Creating one is the step that turns a set of trained weights into something an application can call.
The configuration that matters most is the scaling floor: whether at least one replica is always resident, or whether the endpoint is permitted to drop to zero when idle. That single choice determines both the cost profile and the worst-case latency.
An endpoint is also the natural unit of isolation. Separate endpoints for separate models or environments keep failures, deployments and bills from mixing.
Related terms
Start with the free tier
A magic link creates your account, your tenant and your first API key. No card until you ask for compute.