Embeddings NVIDIAInstall

Summary

Generates text embeddings through NVIDIA’s OpenAI-compatible /embeddings endpoint and exposes them as an n8n AI Embedding output for downstream AI nodes.

Inputs

• Model Name or ID (required) — Select a model from the available models or provide a custom model ID. The default is nvidia/llama-3.2-nemoretriever-1b-vlm-embed-v1.
• Encoding Format — Choose Float (default) or Base64 for the embedding encoding sent in the API request.
• Truncate — Choose whether to leave input untruncated (None, default) or truncate from the Start or End.

Output shape

An AI Embedding connection that supplies embedQuery and embedDocuments functions.

Query text is sent with input type query; document texts are sent with input type passage. The node returns the embedding values from the API response’s data array. It requires NVIDIA Embeddings API credentials and errors if no text is provided or the response lacks data.

Links

• n8n community nodes installation guide