Skip to main content
Deploy a model published through Baseten for Model Labs into your workspace with POST /v1/models. Set source.kind to library_listing and supply the listing identifiers. Baseten creates a model and its first deployment from the listing.

​
Get the listing identifiers

Ask the listing publisher for its lab_display_name and user_defined_listing_id. Your organization must have access to the published listing. Use a Baseten API key with permission to create models in the destination team.

​
Create the model

Replace <LAB_DISPLAY_NAME> and <LISTING_ID> with the identifiers from the publisher. Set source.deployed_model_name to name your model, or omit it to use the listing’s configured name.
The response contains model and deployment objects. Save model.id and deployment.id to check the deployment’s status and call it.

​
Deploy into a specific team

Use GET /v1/teams to find the destination team ID:
Send the same create request body to POST /v1/teams/{team_id}/models, replacing <TEAM_ID> in the URL:
The team ID belongs in the URL. The organization-scoped /v1/models endpoint uses the organization’s default team.

​
Check the deployment

Creation returns before the deployment is ready. Replace <MODEL_ID> and <DEPLOYMENT_ID> with the IDs from the create response and check the deployment status:
When status is ACTIVE, call the deployment with the input format specified by the listing publisher. If the deployment fails, check its deployment logs.

​
Next steps