Skip to main content
Open Playground in the sidebar, pick a model, and send a message. Add up to four models and every message goes to all of them at once, so you can compare answers, speed, and cost on the same prompt.

​
Comparing models

Each column is one model. Click Add model for another column (up to four) and use the picker at the top of a column to change its model at any time. Every column keeps its own conversation: a follow-up replays that model’s own earlier answers, and a column added or swapped mid-conversation only sees the turns it answered itself. Stop cuts every reply short. Under each reply, a stats line shows total time and time to first token, input and output tokens, and the cost of that reply as billed to your organization.

​
System prompt and parameters

Parameters above the columns sets a system message, a temperature (0 to 2), and a maximum reply length for every column. Models that reject sampling controls ignore the temperature; a very low token cap turns a thinking model’s reasoning off so the answer fits.

​
Which models are listed

The picker lists the hosted models Belvedir can serve, grouped by provider, plus the endpoints you registered on the Cloud Inference page. You can also type any other model identifier the router accepts. With Chinese models off under Project Permissions, Chinese-lab models are neither listed nor served. Fine-tuned models run here once you deploy them and register the endpoint on the Cloud Inference page; local models never run here.

​
Billing

A Playground call is billed to your organization per token at the rates on the Pricing page (Inference → Pricing) and appears on the usage page as Playground. Calls to your own registered endpoints run on your provider account and cost nothing on Belvedir. The Playground never routes or pins: the model you pick is the model that answers.