Playground

Test models with the exact same parameters the API accepts — every message is sent to the live backend.

Demo mode — no tokens billed

Parameters

Model

Context
Speed
Input / 1M
Output / 1M

Temperature 0.7

PreciseCreative

Max tokens 1024

System prompt

Chat

Send a message to test the model

Responses stream from the NovaCompute API.

POST /v1/chat/completions · live