google-vertex-ai

Community n8n node for Google Vertex AI (Gemini text, multimodal analysis, Nano Banana image generation/editing, and Veo video). Unofficial, not affiliated with Google.

Package Information

Released: 4 mo ago
Downloads: 55 weekly / 399 monthly
Latest Version: 0.1.3
Author: Akash

Available Nodes

Google Vertex AI icon
Google Vertex AI
Use Google Vertex AI (Gemini) for text, media analysis, and image generation/editing

Documentation

n8n-nodes-google-vertex-ai

An n8n community node for Google Vertex AI. It brings Gemini text, multimodal analysis, image generation/editing (the "Nano Banana" Gemini image models), and Veo video generation into your n8n workflows.

Disclaimer: This is an unofficial, community-built node and is not affiliated with, endorsed by, or sponsored by Google. "Google", "Vertex AI", "Gemini", and related marks are trademarks of Google LLC. You bring your own Google Cloud project and credentials, and your own API usage is billed by Google.

Installation · Credentials · Operations · Usage notes · Compatibility · Development

Installation

From the n8n editor (self-hosted)

  1. Go to Settings → Community Nodes → Install.
  2. Enter n8n-nodes-google-vertex-ai and confirm.
  3. The Google Vertex AI node appears in the node panel.

Manually

# in your n8n instance's user folder (e.g. ~/.n8n)
npm install n8n-nodes-google-vertex-ai

Then restart n8n.

Credentials

Create a Google Vertex AI API credential. It uses a Google Cloud service account:

Field Description
Region Vertex AI location (e.g. us-central1). Use a specific region for image/video; some Gemini models also support global.
Project ID Your GCP project ID where the Vertex AI API is enabled.
Service Account Email The service account's client email.
Private Key The service account's private key (including the BEGIN/END lines).

Setup: create a service account in your GCP project, grant it the Vertex AI User (roles/aiplatform.user) role, enable the Vertex AI API, and create a JSON key — copy client_email, private_key, and your project_id from it. See n8n's Google service account guide. Use the credential's Test button to confirm authentication.

Operations

Text

  • Message a Model — send a prompt (with optional history) to a Gemini model and get a text reply.

Image

  • Analyze Image — send one or more images + a question, get a text answer.
  • Generate an Image — text → image (Nano Banana / Gemini image models). Returns the image as binary.
  • Edit Image — upload one or more images + a prompt, get an edited image back as binary.

Document

  • Analyze Document — send a document (PDF, etc.) and ask questions about it.

Audio

  • Analyze Audio — ask questions about an audio file.
  • Transcribe Audio — transcribe a recording to text.

Video

  • Analyze Video — ask questions about a video.
  • Generate Video — start a Veo text-to-video job; returns an operation name.
  • Get Video Status — check whether a generation operation has finished.
  • Download Video — download the finished video as binary.

Models are selectable per operation: Gemini (text/vision) for analyze/message, the Gemini image models (gemini-2.5-flash-image, gemini-3-pro-image, …) for image generate/edit, and Veo (veo-3.1-generate-001, …) for video.

Usage notes

  • Image input (Analyze / Edit) accepts one or more binary images via the Images field (add as many as you need). MIME type is auto-detected.
  • Image output (Generate / Edit) is written to a binary field (default edited, configurable under Options → Put Output In Field).
  • Video is asynchronous. Wire it as Generate Video → (poll) Get Video Status until done: true → Download Video. Generation takes from tens of seconds to a few minutes. Output is inline base64 → binary video/mp4 (no Cloud Storage bucket required). Veo is region-specific — keep your credential region to a Veo-supported region such as us-central1.

Compatibility

  • Requires a recent n8n with community-node support.
  • Tested against the Gemini, Imagen-family image, and Veo models available on Vertex AI as of mid-2026. Model IDs evolve — if a model returns a 404, pick another from the dropdown or update to a current ID.

Development

npm install
npm run dev      # run n8n locally with this node + hot reload
npm test         # run the jest test suite
npm run lint     # n8n community-node lint
npm run build    # compile to dist/

Releases are published to npm via GitHub Actions with provenance (see .github/workflows/publish.yml); run npm run release to cut a version.

License

MIT

Discussion