Models

Models are the foundation of intelligence in Studio AI. They are responsible for interpreting inputs, generating responses, and performing natural language tasks. Every agent, Skill, or flow configuration depends directly on selecting an appropriate model.

Types of Models Currently, Studio AI works with the following categories:

• Embedded (native): Base models provided by the platform.

  • Cannot be edited or deleted.

  • Serve as starting points for creating custom models.

• Custom: Versions derived from an embedded model, created by the user.

  • Allow customization of name, description, prompt, temperature, and image.

Every custom model uses an embedded model as its base.

Create a Model

1. Go to the side menu and click on “Models.”

2. Choose an embedded model from the list.

3. Click “Customize.”

4. Fill in the following fields:

• Image (optional): Illustrative icon for the model.

• Model name

• Description: For internal identification; this description does not affect the model’s behavior.

• System prompt: Main instructions that define the model’s behavior (e.g., “You are an assistant specialized in answering questions about Studio AI”).

• Temperature: Defines the level of creativity in the responses.

  • Low (e.g., 0.2): More direct and precise answers.

  • High (e.g., 0.8): Freer and more creative answers.

1. Click “Save model.”