NVIDIA NIM Chat Model icon

NVIDIA NIM Chat Model

For advanced usage with an AI chain or AI agent

NVIDIA NIM Chat Model

AI-generated

Summary

Provides a NVIDIA NIM chat model for use as the Chat Model input to n8n AI Agent or AI Chain nodes. Select a model and optionally configure generation, retry, timeout, and response format settings.

Inputs

  • Model (required) — NVIDIA NIM model used for chat completions; the list is loaded from the configured service, with meta/llama-3.1-8b-instruct as the default.
  • Options — Optional settings: frequency and presence penalties (default 0), max retries (2), maximum tokens (unlimited when set to -1), response format (text by default, or JSON), sampling temperature (0.7), timeout in milliseconds (360000), and top P (1). For JSON format, include the word “json” in the chain or agent prompt.

Output shape

A single AI Language Model connection output.

Supplies a ChatOpenAI model configured with the NVIDIA NIM base URL and API key credentials. It does not return chat completion items directly; connect it to an AI Agent or AI Chain. JSON response format is sent as a model response_format setting.

Links

Discussion