Models

Setting Up a Model¶

• Create a new model: On the Models view, click "Add model" to create a new model. You can click on the gear icon, or double click the model's view to access its settings.

• Basic Tab:
  • Name: Enter a unique name for the model.
  • Description: Provide a brief description of the model's purpose or functionality.
  • Model Type: Select the model type (e.g., OpenAI) from the dropdown list.
  • API Key: Input your API key to authenticate requests.
  • Base URL: If the base URL is not one that is pre-filled, you can specify the API base URL (for OpenAI models, it’s usually https://api.openai.com/v1 but for other OpenAI compatible models, this can differ).

• Advanced Tab:
  • Temperature: Adjust the temperature (a value between 0 and 1) to control the creativity of the model's output. Lower values produce more deterministic results, while higher values allow for more diverse responses.
  • Top P: Set the Top P parameter to limit the selection of tokens based on cumulative probability. If you want it unset, leave it as is.
  • Max Tokens: Define the maximum token count for each response (set to "No limit" in the example).
  • Default Headers: Add any additional headers needed for API requests (optional).
  • Tags: Use tags to categorize or label the model (optional but recommended).

• Price Tab:
  • Prompt price per 1K tokens: Enter the price per 1,000 tokens for the model (optional).
  • Completion price per 1K tokens: Enter the price per 1,000 tokens for completions (optional).

• Save the Configuration:
  • After completing the setup in any of the tabs, click Save to confirm and apply the model configuration.

This setup will enable you to link this model to one or more agents for generating responses based on the model's capabilities.