Provider configuration values

Valid provider values for the automation intelligent assistant and when to use each one.

The lightspeed_chatbot_default_provider inventory variable accepts three values. Choose the value that matches your LLM service deployment type, not the API compatibility layer it exposes.

Table 1. Provider values and usage
Provider value LLM service type When to use Example URL pattern
rhoai Red Hat Enterprise Linux AI

Your LLM is deployed on Red Hat Enterprise Linux AI with vLLM Server as the inference engine

 https://<your-rhelai-host>/v1
rhoai Red Hat OpenShift AI Your LLM is deployed on Red Hat OpenShift AI single-model serving platform using vLLM runtime, or your model URL containsrhoai in the hostname https://litellm-*.apps.*.rhoai.*.com/v1
openai OpenAI You are using the OpenAI API platform directly (not a self-hosted model with OpenAI-compatible API) https://api.openai.com/v1
azure Microsoft Azure OpenAI Your LLM is deployed on Microsoft Azure OpenAI service https://<your-resource>.openai.azure.com/openai/v1
Note

The most common error is setting lightspeed_chatbot_default_provider="openai" when the model is actually deployed on Red Hat OpenShift AI with an OpenAI-compatible API endpoint. In this scenario, even though the API is OpenAI-compatible, the correct provider value is rhoai because that is the actual deployment platform.