NVIDIA in n8n: Nemotron Chat Model vs NVIDIA NIM Community Nodes

n8n has one built-in NVIDIA node: the NVIDIA Nemotron Chat Model, a chat model for AI agents and chains that works with Nemotron models on build.nvidia.com and with a self-hosted NVIDIA NIM. For other models from NVIDIA’s catalog, embeddings, speech recognition or image analysis, n8n users install community nodes. This guide covers the built-in node and its credential (from n8n’s documentation) and the NVIDIA community packages on our hub (from the descriptions and READMEs their authors publish on npm, and our own download data).

Quick comparison

The NVIDIA Nemotron Chat Model

You pick a Model; n8n loads the Nemotron models from the endpoint in your credential and, if it can’t reach it, falls back to a list of well-known Nemotron model IDs. Options are Frequency Penalty, Maximum Number of Tokens (-1 for the model default), Response Format (Text or JSON; with JSON, put the word json in your prompt), Presence Penalty, Sampling Temperature, Timeout, Max Retries and Top P.

The NVIDIA credential

One credential covers both setups. It takes a Base URL, by default https://integrate.api.nvidia.com/v1 for build.nvidia.com, and an API Key:

• build.nvidia.com: sign in, open a Nemotron model in the catalog, select Get API Key and paste the key into n8n.
• Self-hosted NIM: set the Base URL to your NIM endpoint including /v1, for example http://localhost:8000/v1, and leave the API key blank if your NIM doesn’t require authentication.

NVIDIA community nodes

Our NVIDIA hub lists 10 packages that pass our quality rule, with 1,800 npm downloads a week between them. The most downloaded, with the description each author gives on npm (weekly npm downloads as of 28 October 2026):

Versions and dates come from the npm registry. Downloads are spread evenly across the top seven packages.

Chat models for any NIM model

Several packages do what the built-in node does, but for every model on the endpoint rather than Nemotron only: n8n-nodes-nvidia-nim-chat and n8n-nodes-nvidia-nim-gabriel both connect to AI agents and chains as a language model, load the model list from your account or self-hosted NIM, and take a base URL, so they work with build.nvidia.com and your own NIM. n8n-nodes-nvidia-chat-modell, with a README in Portuguese, is built for testing and comparing models and switches between NVIDIA NIM, Google Gemini and Anthropic Claude.

Chat as a regular node, and image analysis

n8n-nodes-nvidia-nim works as a regular workflow node rather than an agent sub-node, with a chat node and an NVIDIA NIM Image Analysis node for vision models that takes an image as a URL, base64 or data URL plus a prompt. Its README lists n8n 1.0.0 or newer and an NVIDIA NGC API key. n8n-nodes-nvidia-kimik2 is also a regular node, for OpenAI-compatible chat completions with streaming and a thinking mode.

Embeddings, speech and reranking

• @sateeshreddy/n8n-nodes-nvidia-nim: two nodes, a language model for AI agents and chains and an embeddings node that plugs into n8n’s vector stores such as Qdrant, Pinecone, PGVector, Supabase, Weaviate and Chroma.
• n8n-nodes-nvidia-nim-whisper-v2: speech to text and translation with Whisper Large V3 through NVIDIA Riva, for WAV (mono 16-bit), OPUS and FLAC audio, with automatic language detection; works with NVIDIA’s cloud API and self-hosted NIM.
• n8n-nodes-reranker-nvidia and n8n-nodes-embedings-nvidia: reorder documents retrieved from a vector store, and create text and image embeddings; both have short READMEs and date from October 2025.

Check each node page here for the latest release and, where it has one, its GitHub repository for open issues before you build on a package.

How to choose

• A Nemotron model in an AI agent: the built-in NVIDIA Nemotron Chat Model.
• Another model from NVIDIA’s catalog in an AI agent: n8n-nodes-nvidia-nim-chat or n8n-nodes-nvidia-nim-gabriel.
• NVIDIA embeddings for a vector store: @sateeshreddy/n8n-nodes-nvidia-nim, for example with the Qdrant Vector Store.
• Transcribe audio: n8n-nodes-nvidia-nim-whisper-v2.
• Describe images: the image analysis node in n8n-nodes-nvidia-nim.

To install a package on a self-hosted instance, an owner or admin enters its npm name in Settings > Community Nodes; full steps are in How to Install n8n Community Nodes. On n8n Cloud only verified community nodes can be installed, see Verified vs Unverified n8n Community Nodes. Similar guides: OpenAI and OpenRouter.

FAQ

Does n8n have an NVIDIA node?

Yes, the NVIDIA Nemotron Chat Model, a chat model sub-node for AI agents and chains. Other NVIDIA models and services come from community nodes.

Where do I get an NVIDIA API key for n8n?

On build.nvidia.com: sign in, open a Nemotron model in the catalog and select Get API Key. A self-hosted NIM that doesn’t enforce authentication needs no key.

Can I use a self-hosted NVIDIA NIM with n8n?

Yes. In the NVIDIA credential, set the Base URL to your NIM endpoint including /v1, such as http://localhost:8000/v1.

What is the most used NVIDIA community node?

n8n-nodes-nvidia-nim, with 292 weekly npm downloads on 28 October 2026, closely followed by several other NIM chat packages.