Endpoint
The URL or address where you send a request and the model sends back its answer.
Think of it like
Like a ticket window at the post office — you hand in your request and wait for the response.
Example
You call `https://api.example.com/v1/completions` with your prompt, and the endpoint replies with generated text.
How it actually works
An endpoint is a running service listening on a network address. It receives JSON payloads, passes them through the model, and streams or returns the output. Uptime, latency, and throughput all depend on how the endpoint is built and where it runs.
For product teams
Endpoints are the interface between your app and the model — their reliability and speed shape user experience.
For engineers
An HTTP(S) endpoint backed by a running inference server that decodes requests, invokes the model, and serializes responses.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome