Make Predictions

This page will show you how to make predictions using your deployments.

Use the ModelZ SDK

The easiest way to make predictions is to use the ModelZ SDK.

pip install modelz-py

For example, if you have deployed a Deployment a2a4bb7d-2899-4472-a3b4-ee69f04650a9 and you want to make a prediction using the ModelZ SDK, you could run the following code:

import modelz
 
 
APIKey = "mzi-abcdefg..."
 
cli = modelz.ModelzClient(key=APIKey, deployment="a2a4bb7d-2899-4472-a3b4-ee69f04650a9", timeout=10)
cli.inference(params={})

Use the ModelZ CLI

modelz inference --help

Use the curl command

You could also use the curl command to make predictions.

For example, if you have deployed a Deployment a2a4bb7d-2899-4472-a3b4-ee69f04650a9 and you want to make a prediction using the curl command, you could run the following command:

curl --location 'https://a2a4bb7d-2899-4472-a3b4-ee69f04650a9.modelz.io/inference' \
--header 'Content-Type: application/json' \
--header 'X-API-Key: mzi-abcdefg...' \
--data '{
    ...
}'