Playground
Chat with your deployed models
Send prompts to a model you deployed. Each message calls your endpoint and bills credits per request.
This calls your real deployment
Pick one of your live deployments and chat with it. Responses stream from the machine hosting it (your fleet or on-demand), and each request meters usage against your wallet credits.
How it works →No live models yet
Deploy an open-source model to get an endpoint you can chat with here.