Cloudera Docs

Registering Models

This section details the procedure for managing Large Language Models (LLMs) within Agent Studio, covering how to register new models, the supported model providers, and how to edit and test already registered models for proper functioning.

Registering Models

To register a new model, click the Register New Model button, under the LLMs tab. The Register Model window is displayed, and enter the following details:
  1. Model Provider - The following model providers are currently supported:
    • Cloudera AI Inference
    • OpenAI
    • Azure OpenAI
    • Google Gemini
    • Anthropic
    • AWS Bedrock
    • OpenAI compatible providers - The OpenAI compatible providers option allows you to connect to any model endpoint that adheres to the OpenAI API specifications.
  2. Model Alias - Enter a unique name for your model to be referenced across the studio.
  3. Model Identifier - Enter the model identifier depending on the Model Provider that you choose.
  4. API Base - Enter the base URL for the model's API.
  5. API Key - Enter the API key for accessing the model's service
  6. Toggle the Set as Default button to make this as the default model for the studio.

Editing registered models

A list of all the registered models are displayed in the Models window.

To update the details of a registered model, click the Edit Model button under the Actions menu. All the fields can be edited except the model provider. To replace the API key, enter the new key in the API Key field; leaving the field blank will keep the existing API key unchanged.

Testing models

In Agent Studio, you can validate a registered model's credentials and network connectivity by testing the model with a simple prompt.