Setting Up an LLM Connection for Copilot

Copilot requires a large language model (LLM) to work. The LLM allows Copilot to interpret natural-language questions, generate answers to queries, and return results in plain language. The steps below explain how to set up an LLM connection for Copilot via the Administration application.

You can elso establish an LLM connection for Copilot via the Create Prompt dialog.

1. In the Administration application, select Connections > LLM > Connections.
2. Click Add LLM Connection > Create New LLM Connection. The Create LLM Connection dialog displays.

4. Provide the necessary information using the table below as a guide. Your LLM administrator will be able to provide you with these details.

Setting	Description
Title	Required field that specifies the unique name for the annotator.
LLM Endpoint	Required field that specifies the URL of the LLM service.
LLM Provider Name
Required field that specifies the name of the LLM provider.

Some examples of LLM providers compatible with Graph Studio include:

• OpenAI
• Azure OpenAI
• Gemini
• Anthropic

| | LLM Model Name | The name of the model to use for LLM operations. | | LLM API Key | The authentication token granting access to the LLM service. This same key should be provided in the Confirm LLM API Key field. | | Can Use LLM Connection | Specifies the roles granted permission to use the LLM connection. |

6. Click Confirm to save your connection.

The new LLM connection is added to the LLM Connections page.