› Integration › Generative AI › Configure Tungsten AI provider

Configure Tungsten AI provider

The "Tungsten" AI provider is available by default in TotalAgility.

It is used for "Copilot for development" and "Copilot for extraction." Additionally, the Tungsten provider supports images.

Request a database script from Technical Support to change the Tungsten AI provider URLs in TotalAgility when acessing Tungsten AI provider through a proxy server that requires different URLs.

On upgrading, the Tungsten AI provider is available with the following exceptions:

  • If no provider was previously set to "Use for Development" or "Use for Extraction," the Tungsten AI provider has both options set to true.
  • If a provider was previously set to "Use for Development," the Tungsten AI provider has "Use for Development" set to false, while the existing provider configuration remains unchanged.

  • If a provider was previously set to "Use for Extraction," the Tungsten AI provider has "Use for Extraction" set to false, while the existing provider configuration remains unchanged.

Procedure
  1. Navigate to Integration > Generative AI.

    The Generative AI providers page appears.

  2. Select Tungsten AI provider from the list.
  3. Configure the following settings:

    Configuration settings

    Description

    ID

    Read-only field.

    Display name

    Read-only field.

    Use provider for Copilot for development

    The AI provider is used as the Copilot for development. (Default: Selected)

    Default provider for Copilot for extraction

    The AI provider is used as an Extraction provider. (Default: Selected)

    Default provider for classification

    The AI provider is used for classification. (Default: Selected)

    Default provider for extraction

    The AI provider is used for extraction. (Default: Selected)

    Standard/Vision Model
    Temperature

    Set the process temperature for the AI provider to control the level of creativity in the AI-generated text. (Default: 0.5, Minimum value: 0, and Maximum value: 2)

    When the AI provider is configured as "Copilot for extraction", the system ignores the temperature parameter and always uses a fixed value of 0. This ensures deterministic, repeatable extraction results; therefore, any change in temperature does not affect the extraction output.

    Reasoning effort

    The reasoning effort parameter controls how much the model thinks before it answers. Optional. Select the reasoning effort level for reasoning models.

    Option Description
    Auto (Default) System automatically determines the reasoning effort level.
    None Use for latency-critical work.
    Low Use for efficient reasoning with tool use, planning, and search.
    Medium Use as a balanced starting point for most workloads.
    High Use for complex debugging and deep planning where quality matters.
    Extra High Use for deep research and agentic tasks with long runs.

    • For the Tungsten AI provider:
      • During process or form creation using Copilot for development and MCP server interactions, the reasoning effort parameter is not applicable because these operations rely on functions and tools.
      • The Tungsten AI provider does not allow the reasoning effort parameter to be used together with functions or tools. As a result, the parameter is automatically excluded when using functions or tools.
    • For non-reasoning models, use "Auto".
    Timeout in seconds

    Configure the timeout period in seconds for the provider so the request does not become unresponsive indefinitely. (Default: 300, minimum: 120, and maximum: 3600)

    Retry count

    Set the retry count for the Tungsten AI provider. (Default: 5, minimum: 0, and maximum: 100)

    Vision timeout in seconds

    Configure the vision timeout period in seconds for the provider. (Default: 300, minimum: 120, and maximum: 3600)

    Vision retry count

    Set the vision retry count for the provider. (Default: 5, minimum: 0, and maximum: 100)

    Token limit

    By default, the token limit for the Tungsten AI provider is 3000.

    Authentication
    OAuth server

    Optional. Use this setting only when accessing the Tungsten AI provider through a proxy server that requires OAuth authentication. The system automatically retrieves and refreshes access tokens for API calls.

    See Configure OAuth server.

    Custom headers

    Optional. Use this setting only when accessing the Tungsten AI provider through a proxy server that requires specific custom headers. Configure custom headers to pass authentication information or client identifiers.

    If using OAuth server type custom headers, an OAuth server must be selected.
    1. Select Add.
    2. Enter the header Name to be sent in the HTTP request.
    3. Select the Type:
      • Standard (Default): Enter the Value (passed at runtime as a static value).
      • OAuth server: The access token is automatically retrieved and used.
    4. Select Add.
    5. Repeat the above steps to add additional custom headers.

    The Temperature, Timeout in seconds, Retry count, Vision timeout in seconds, and Vision retry count configuration settings are not applicable when the Tungsten AI provider is used as Copilot for extraction.

  4. Select Save.

At runtime, when using Copilot for development, Copilot for classification, or Copilot for extraction, the system automatically:

  • Retrieves access tokens from the OAuth server (if configured).
  • Includes the access token in the Authorization header.
  • Passes the specified custom headers with every API call to the Tungsten AI provider.

More info