Deploying customized models

After customizing and evaluating your model, deploy it for inference. The deployment path depends on the model type. For end-to-end serverless deployment workflows (including one-click deploy from Studio UI), see Serverless model customization.

Amazon Nova models

Customized Amazon Nova models deploy to Amazon Bedrock. After your training job completes, the model is registered in SageMaker AI Model Registry and can be deployed as a Bedrock imported model.

For deployment instructions, see Importing custom models in Amazon Bedrock.

Open weight models (OSS)

Customized open weight models deploy to SageMaker AI Inference endpoints.

Studio UI

Navigate to Models → My models → select your logged model → Deploy.