Is there a Google Gemini node in n8n?

Yes. The Google Gemini node handles text, images, audio, video, documents, media files and file search stores. There are also Gemini chat model and embeddings sub-nodes for AI Agent workflows.

Which credential does it use?

Google Gemini(PaLM) API with a Gemini API key from Google AI Studio. On n8n Cloud you can use Gateway credits instead.

How do I replace Generate Embeddings?

Use the Embeddings Google Gemini sub-node inside a vector store flow, or an HTTP Request to the embedContent endpoint with the same credential.

Do Gemini steps convert automatically?

Not yet. Gemini actions arrive as marked placeholders with the model and prompt; each is one Google Gemini node. Code steps that call the Gemini REST API convert automatically.

Pipedream Google Gemini steps in n8n

Updated 2026-10-21 · 3 min read · n8n: Google Gemini (Native n8n node)

Pipedream's Google Gemini app has three actions: generate content from text, generate content from text and an image, and generate embeddings. n8n's native Google Gemini node covers all of that and more, and the credential is a single API key.

The n8n Google Gemini node

n8n-nodes-langchain.googleGemini:

Resource	Operations
Text	Message a Model
Image	Analyze Image, Generate an Image, Edit Image
Document	Analyze Document
Audio	Analyze Audio, Transcribe a Recording
Video	Analyze Video, Generate a Video, Download Video
Media File	Upload Media File
File Search	Create, Delete and List File Search Stores, Upload to File Search Store

For agents and RAG, n8n also has the Google Gemini Chat Model and Embeddings Google Gemini sub-nodes. There is no Gemini trigger, and Pipedream's Gemini app has no triggers either.

What maps where

Pipedream (app `google_gemini`)	In n8n	Status
Generate Content from Text	Google Gemini → Text → Message a Model	manual, 1 node
Generate Content from Text and Image	Image → Analyze Image (image as URL or binary)	manual
Generate Embeddings	Embeddings Google Gemini sub-node, or HTTP Request to `embedContent`	manual
Code step calling `generativelanguage.googleapis.com` with `this.google_gemini.$auth.api_key`	Code node; key becomes `GOOGLE_GEMINI_API_KEY`	automatic
Code step importing `@google/generative-ai` or `@google/genai`	Code node; SDK listed as npm package	automatic, review

Gemini actions are not in the converter's automatic list, so they arrive as placeholders with the model ID and prompt text in a comment.

Credentials

Create a key at Google AI Studio (aistudio.google.com/apikey).
n8n: Credentials → New → Google Gemini(PaLM) API, paste the key. The host must stay https://generativelanguage.googleapis.com; custom hosts and proxies are not supported.
On n8n Cloud, "Use Gateway credits" is an alternative to your own key.

This is the Gemini Developer API key, not a Google Cloud Vertex AI service account. If the Pipedream workflow used Vertex AI through code, that is a different setup.

Typical migration: describe product photos

Pipedream: Google Drive New Files → Download File → Generate Content from Text and Image ("write an alt text") → Google Sheets Add Single Row. n8n: Google Drive Trigger → Google Drive → Download → Google Gemini → Image → Analyze Image (binary input) → HTTP Request (Sheets append).

Pitfalls

Quotas are per project, not per key. Gemini limits requests per minute, tokens per minute and requests per day per Google Cloud project; the daily count resets at midnight Pacific time. Two keys in one project share the quota.
Tiers. Higher limits depend on billing history (for example Tier 2 after $100 spend and 3 days). A new key on a fresh project starts low; plan the first bulk run accordingly.
Loops. n8n calls the node once per item. 500 rows means 500 requests; add a Wait node or batch the prompt.
Output shape. The node's output is not the raw candidates[0].content.parts structure that Pipedream returned. Re-check expressions downstream after one test run.
SDK imports do not work in Code nodes on n8n Cloud. Use the Gemini node or HTTP Request instead.

How to move your Gemini workflows

Export the workflows from Pipedream.
Upload on /convert; the free report lists Gemini steps and the key variable.
Download the n8n JSON: 1 workflow is free; full export: $3 per workflow, $19 minimum — coming soon.

More in the Pipedream to n8n guide; test SDK-based code in the code step checker. Related: OpenAI, Anthropic, Google Sheets.

FAQ

Is there a Google Gemini node in n8n?: Yes. The Google Gemini node handles text, images, audio, video, documents, media files and file search stores. There are also Gemini chat model and embeddings sub-nodes for AI Agent workflows.
Which credential does it use?: Google Gemini(PaLM) API with a Gemini API key from Google AI Studio. On n8n Cloud you can use Gateway credits instead.
How do I replace Generate Embeddings?: Use the Embeddings Google Gemini sub-node inside a vector store flow, or an HTTP Request to the embedContent endpoint with the same credential.
Do Gemini steps convert automatically?: Not yet. Gemini actions arrive as marked placeholders with the model and prompt; each is one Google Gemini node. Code steps that call the Gemini REST API convert automatically.