Create Workloads¶
Premium feature
The Workload API is a premium feature. Contact your DataRobot representative or administrator for information on enabling the feature.
A Workload is the stable identity that holds your container—an invoke URL, a sharing model, an importance level, and a lifetime policy. This section covers the object model, the draft-vs-locked decision, an end-to-end walkthrough, and the REST and console surfaces for creating and reading Workloads.
| Resource | Description |
|---|---|
| Workload concepts | Describes the object model, full lifecycle state reference, protons, lifetime policies, the draft vs. locked decision, importance, and sharing. |
| Tutorial: Deploy a production-ready container | Deploy a locked artifact with importance, sharing, and monitoring using cURL, the CLI, or Pulumi. A downloadable notebook version is linked from the page. |
| Deploy an open-weight model from Hugging Face | Walks through sizing a GPU compute bundle and serving a Hugging Face Hub model with vLLM's OpenAI-compatible server. |
| REST: Create and read Workloads | Describes POST / GET Workloads with links to the API reference. |
| UI: View deployed Workloads | Learn about the Console list, filters, and Workload-centric visibility. |