Configuring LLMs

link

• Replace UiPath LLM subscription: Replace UiPath-managed subscriptions with your own, provided they match the same model family and version already supported by the UiPath product. This allows for seamless swapping of UiPath-managed models with your subscribed models.
• Add your own LLM: Use any LLM that meets the product's compatibility criteria. To ensure smooth integration, your chosen LLM must pass a series of tests initiated through a probe call before it can be used within the UiPath ecosystem.

• Compatibility requirements: Your chosen LLM or subscription must align with the model family and version supported by the UiPath product.
• Setup: Make sure you properly configure and maintain all required LLMs in the custom setup. If any component is missing, outdated, or incorrectly configured, your custom setup may cease to function. In such cases, the system will automatically revert to a UiPath-managed LLM to ensure continuity of service, unless UiPath LLMs are turned off through an Automation Ops policy.
• Cost-saving: If your custom LLM setup is complete, correct, and meets all necessary requirements, you may be eligible for a Reduced Consumption Rate.

• Unified Pricing: Enterprise Platform, Standard Platform, Basic Platform, App Test Platform Enterprise, App Test Platform Standard.
• Flex: Advanced Platform, Flex Standard Platform.

Direct connectivity through Integration Service​

Private network connectivity through Relay​

• Azure Open AI
• Open AI
• Amazon Web Services – Use this connector to configure Anthropic Claude models.
• Google Vertex
• Open AI V1 Compliant LLM – Use this option to connect to any LLM provider whose API follows the OpenAI V1 standard. For details, refer to the OpenAI V1 Compliant LLM connector documentation.
  Note:

  To use Integration Service connections, you must add the Integration Service outbound IP ranges to your allowlist.

1. Create the Integration Service connection​

1. In Integration Service, create a connection to your LLM provider.
2. Choose the folder where the connection will be stored.
3. Complete the authentication as required by the selected connector.
   Note:

   The folder you choose controls both security and visibility. To prevent unauthorized access, create the Integration Service connection in a private, non-shared folder. However, note that model visibility is determined by access to this folder. If an admin does not have access to it, the associated model configuration will not appear in the list.

2. Add a new LLM configuration​

1. Navigate to Admin > AI Trust Layer > LLM configurations.
2. Select the Tenant.
3. Select Add configuration.
4. Choose the Product (for example, Agents) and Feature (for example, Design, Evaluate & Deploy).
5. Select the Connections Folder.

3. Configure the model​

• LLM Name – This field supports two configuration options, depending on your use case:
  • Select a model from the list – Choosing a model from the predefined list replaces the UiPath-managed LLM subscription with your own subscription for that same model. This scenario is referred to as Bring Your Own Subscription (BYOS).
  • Add custom alias – Entering a custom name lets you configure a model that is not included in the predefined list of recommended models for that product. This scenario is referred to as Bring Your Own Model (BYOM).
    Note:

    The Add a custom alias option is available only for products that support custom models. To learn which products support custom models, refer to the table below.

• API Type – The API endpoint supported by the LLM (for example, Open AI Chat Completions). This must match the endpoint exposed by your provider.
• Connector – The Integration Service connector (for example, Microsoft Azure OpenAI, Amazon Web Services).
• Connection – The specific Integration Service connection created earlier. If no connection is available, create one in Integration Service.
• LLM identifier – The model identifier exactly as it appears in your LLM subscription.
  • For Azure-hosted models: enter the model deployment name/identifier.
  • For AWS Bedrock cross-region inference: enter the inference profile ID. You must prepend the model ID with your region code, for example: eu.anthropic.claude-3-5-sonnet-20240620-v1:0.
  • For other providers: use the model name as defined in your subscription.

4. Validate and save​

1. Select Test configuration to verify that the endpoint is reachable.

   • The platform validates the connectivity.
   • Ensuring the correct model is configured remains your responsibility.

2. If the validation is successful, select Save to activate the configuration.

   

• Check status – Verify the status of your Integration Service connection. This action ensures that the connection is active and functioning correctly.
• Edit – Modify any parameters of your existing connection.
• Disable – Temporarily suspend the connection. When disabled, the connection remains visible in your list but doesn't route any calls. You can re-enable the connection when needed.
• Delete – Permanently remove the connection from your system. This action disables the connection and removes it from your list.

• For features with a selectable model — where you choose which model to use — you can configure one or more models; unconfigured models continue to use UiPath-managed subscriptions.
• For features with a fixed model set — where the feature uses a predetermined set of models — all models must be configured for the feature to work; partial configuration is not valid.

• Ensure your LLM supports:
  • Tool (function) calling – The model must be able to call tools or functions during execution.
  • Disabling parallel tool calls – If supported by your provider, the model should offer the option to disable parallel tool calls.
• When using custom models, Agents default to a 4096 token limit regardless of the model’s true capacity, since UiPath cannot infer token limits for customer-defined deployments.