Azure DeepSeek LLM icon

Azure DeepSeek LLM

Interact with DeepSeek LLM on Azure AI Foundry

Actions2

Chat Completion

AI-generated

Overview

This node interacts with the Azure DeepSeek Large Language Model (LLM) to generate chat completions based on user and system prompts. It is useful for scenarios where conversational AI responses are needed, such as virtual assistants, customer support bots, or interactive chat applications. Users provide a system prompt to set the assistant's behavior and a user prompt to generate a relevant response from the model.

Use Case Examples

  1. A customer support bot that uses the system prompt to define its role and the user prompt to answer customer queries.
  2. An interactive assistant that helps users with specific tasks by generating responses based on the conversation context.

Properties

Name Meaning
System Prompt The system message that sets the behavior of the assistant, guiding how it responds.
User Prompt The prompt from the user to send to the model, which the model uses to generate a response.
Additional Options Optional parameters to customize the model's response, including temperature, max tokens, top P, frequency penalty, presence penalty, streaming response, and request timeout.

Output

JSON

  • content - The generated text content from the model's response.
  • _fullResponse - The full raw response object returned by the Azure DeepSeek LLM API.

Dependencies

  • Requires an Azure DeepSeek API key credential for authentication.

Troubleshooting

  • Timeout errors may occur if the request takes too long; increase the request timeout setting to resolve.
  • Network errors may occur if no response is received; check network connection and Azure endpoint configuration.
  • API errors from Azure DeepSeek LLM include error messages returned by the service; review error details for troubleshooting.

Links

Discussion