Skip to main content
This feature may not be available depending on the terms of your contract. Please contact your Omni Account Executive or Omni support to be connected to them if you are interested in this feature. Once enabled, Organization Admin permissions are required to access and modify AI model settings.
By default, Omni uses AWS Bedrock and you can control the default Claude model used for various AI features by adjusting the ai_settings parameter in the Omni model. If you prefer to use a different provider or your own API key, Omni also supports the following providers: Anthropic Direct, Google Vertex AI, OpenAI, Snowflake Cortex, Databricks Genie, or Grok (xAI). Provider settings are configured in AI Hub > General > Model.

​
Supported providers

AWS Bedrock

Default provider. Uses Omni’s managed AWS credentials and model selection.

Anthropic Direct

Direct API access to Anthropic’s Claude models. Allows you to select which model is used.

Google Vertex AI

Access to Claude models through Google Vertex AI using workload identity federation (WIF). Requires configuration in GCP.

OpenAI

Access to GPT models. Supports integration with OpenAI-compatible APIs like Azure OpenAI and Ollama and custom model identifiers for bring-your-own-model (BYOM) configurations.

Snowflake Cortex

Access to Anthropic’s Claude models through Snowflake Cortex. Requires configuration in Snowflake.

Databricks Genie

Access to models via Databricks serving endpoints. Requires configuration in Databricks.

Grok (xAI)

Access to xAI’s Grok models.

​
Configuration options

Omni has not thoroughly tested all model options. Performance may vary depending on the model selected.
Omni uses a three-tier model system to balance performance and cost across different AI features:
  • Smartest model - The most capable model, used for the most complex reasoning tasks requiring advanced capabilities.
  • Standard model - A balanced model for typical query generation and detailed questions.
  • Fastest model - An optimized model for simpler, high-volume tasks like text completion and basic assistance.

​
Available settings

  • Provider - The AI model provider to use for your organization. The provider determines which models are available and how API requests are handled.
  • Smartest model - The most capable model in your provider’s lineup, used for complex reasoning tasks. Not available for AWS Bedrock or Databricks Genie, which use region-based model selection.
  • Standard model - The model used for typical query generation and analytical tasks. Choose a balanced model for good performance across use cases.
  • Fastest model - The model used for simpler, high-volume tasks like text completion and basic assistance. A smaller, faster model is recommended to optimize performance and cost.
  • Custom model - Applicable only to OpenAI. Override the default model selection with a custom model ID. See Custom model identifiers with OpenAI for more information.
  • API key - Your API key for the selected provider. API keys are securely stored and not displayed after saving. Required for:
    • Anthropic Direct
    • Google Vertex AI - Uses workload identity federation (WIF) credentials instead of an API key. Requires GCP project, region, WIF provider audience, and optional service account impersonation URL
    • OpenAI - Applicable to built-in models and custom models
    • Snowflake Cortex - Use a Snowflake Programmatic Access Token (PAT)
    • Databricks Genie - Use a Databricks Personal Access Token (PAT)
    • Grok
  • Base URL - The base URL for API requests. Available for:
    • OpenAI - Use this to connect to OpenAI-compatible APIs like Azure OpenAI or self-hosted Ollama instances
    • Snowflake Cortex - Provide your Snowflake account’s base URL
    • Databricks Genie - Provide your Databricks workspace URL (e.g., https://your-workspace.cloud.databricks.com)