Model Connectors
Connection Types
- Managed: Configured by the sales team, which includes token grants as part of contractual agreements. Please contact sales for managed connection setups.
- Custom Connections: Can be set up by users directly under specific plans without the need for token grants. Custom connections allow greater flexibility but require a special plan.
Custom Connections
Model Configuration
This dialog allows for the configuration of key parameters necessary for model operation:
- Name: Allows users to set or update the model's name for identification within the platform.
- Provider: Users can select the AI service provider that will dictate available configuration options and operational settings. The supported providers are:
- OpenAI: Standard configurations for OpenAI models.
- Azure OpenAI: Provides additional settings such as endpoint specifications and token limits.
Advanced Settings
Specific settings are available based on the selected provider:
- Endpoint (Azure OpenAI only): Specify the endpoint URL for Azure OpenAI services.
- Temperature (Azure OpenAI only): Configure the randomness or creativity level of the AI's responses.
- Max Tokens (Azure OpenAI only): Define the maximum number of tokens the AI model can generate in a single operation. This setting is exclusive to Azure OpenAI due to its specific API requirements.
Credentials Management
- Credential Selection: Attach existing credentials for API authentication. Credentials management details can be accessed via the Credentials Management Documentation.