An endpoint is a server with preinstalled software that automatically downloads neural network weights and launches an OpenAI-compatible API for making requests to the server.
A private endpoint is created within a user's project, belongs exclusively to that user, and is billed per server configuration according to our infrastructure pricing.
Public endpoints cannot be created by users — they are provisioned by the immers.cloud technical team. They are shared among multiple users via a unified API. To access the chat.immers.cloud load balancer API, you need an access token and a positive account balance.